Conceptual Design of an "Inexpensive" Single-Seat Motorglider

Discussion in 'Member Project Logs' started by Topaz, Sep 10, 2014.

Help Support HomeBuiltAirplanes Forum by donating:

  1. Aug 28, 2015 #41

    Topaz

    Topaz

    Topaz

    Super Moderator Staff Member

    Joined:
    Jul 30, 2005
    Messages:
    13,727
    Likes Received:
    5,407
    Location:
    Orange County, California
    Some Quick Trade Studies
    I’ve been meaning to get these published here, but “life” keeps getting in the way. Sorry for slow pace. They were done days ago.

    Trade studies are a way of saying, “what if” with the design and determining an answer. With the engine question relatively put to rest, I wanted to look at the airplane as it sits right now and see “what if” about certain performance parameters, and how they affected that engine selection process. Why is that important to me? The engine is probably the single largest contributor to the total cost of the project. It’s certainly the single most costly component that I would purchase. Nailing down the engine chosen versus the requirements will be fixing a large percentage of the total project cost right now. As I mentioned in the discussion thread for this project, I’m absolutely serious about this being a “design to cost” effort. Trade studies are a way of cross-checking that I’m doing the right thing.

    In terms of creating these graphs, there’s nothing terribly sophisticated going on here - I just hooked my sizing-related spreadsheets together (result linked below) and varied the chosen parameter over a range, one at a time, and then plotted the results in a separate spreadsheet. Here we go:

    W[SUB]0[/SUB] versus BSFC
    One of the possible engine options I found is the 28hp Hirth F-33. It’s light, at first glance it’s powerful enough, and it’s in the price range. But it has a BSFC of 0.76, which is typical of two-stroke engines like this. The airplane was size using an assumed BSFC of 0.43. What effect does the increased fuel consumption have on the final size of the aircraft?

    W0-versus-BSFC.jpg

    As it turns out, the design is fairly sensitive to BSFC. Installing the F-33 instead of an equivalent four-stroke engine will increase the weight of the entire aircraft to about 810 lbs., up from 724 lbs. That's because the F-33 consumes a lot more fuel per horsepower per unit time, and that extra fuel weighs more, and it requires more airplane to carry it. That larger airplane weight, when you back out through the 25hp/lb. power loading, means the engine will have to put out somewhere in the range of 32-33hp for the aircraft to meet all the performance requirements. The F-33 is 28hp, and so does not meet the need here because of its high BSFC.

    W[SUB]0[/SUB] versus Range
    What about range? I’ve set a rather short range requirement for this airplane, since it’s not intended (or likely to be used) for long-range cross country flights. But what price would I pay for a longer range? Would the extra fuel required (and the extra airframe required to carry that extra fuel) bump the engine required to a larger (and more expensive) size? Let’s see:

    W0-versus-Range.jpg

    Remarkably, the sized design is not terribly sensitive to range. Increasing the range from 350 sm to 500 sm (including reserves) only increases W[SUB]0[/SUB] by about 50 pounds. With my chosen power loading, that keeps the required engine power well within the capabilities of the Hummel ½ VW I’ve baselined. Very interesting. This is probably due to the long wings and relatively high cruise L/D value. Before I finish this initial sizing process, I may consider whether or not I could actually take advantage of a range increase in the real world, or if I should continue with my original requirement.

    W[SUB]0[/SUB] versus Cruise L/D
    This trade study stems from the uncertainty I have in my cruise L/D calculation. I predicted a value of 11 but does it really matter if I’m off a couple of integers or not? Let’s see:

    W0-versus-L-D.jpg

    And the answer is: “Not much”. Engine power required is relatively stable across a range of cruise L/D values of 9-14, and my airplane is very likely to fall into that range somewhere.

    Doing more trade studies at this point would be somewhat counter-productive, because this first-order sizing method is not terribly accurate. It’s a way to get the design “into the ballpark”, not a method for final design. These provide good information that is very useful for me now. When I do the final sizing and optimization, I’ll be able to look at these trades and others in much more useful and accurate ways.

    Complete Initial Sizing Spreadsheet: https://app.box.com/s/ric1gks0y7ay0rn1qf04o9ykxjov328g
     
    Last edited: Aug 29, 2015
  2. Sep 9, 2015 #42

    Topaz

    Topaz

    Topaz

    Super Moderator Staff Member

    Joined:
    Jul 30, 2005
    Messages:
    13,727
    Likes Received:
    5,407
    Location:
    Orange County, California
    Review and Finalizing Initial Sizing
    The first, basic, sizing of the airplane is done, but I want to do a quick review to see if everything is still exactly what I want, based on the work so far.

    Total Cost of Project
    When I wrote up the specifications for this project, I was making the assumption that I would be tying the airplane down at Skylark Field (CA89), since tie-downs are extremely inexpensive there. I had a desire for trailerability, but no real need. In the $8,500 threshold cost for the airplane, I casually said that the trailer would be included in that cost, if I needed one.

    That’s awfully vague, and leaves too much “unknown” for me to really judge whether the project has succeeded in meeting the cost goal. For example, it unfairly penalizes the airplane if a trailer is needed. I really didn’t think this through very well!

    However, now’s the time to make the change, so I’m revising the specification sheet to make the threshold price $8,500 not including trailer. If this project moves forward to being built, and I need a trailer, I’ll scrape together the money for it somehow.

    Rate of Climb
    Over in the discussion page, we talked about climb rate. So far, my specification has to been to meet the FAR 23.65 and CS 22.65 minimum climb requirements at my “hot and high” density altitude condition of 95°F at Hemet-Ryan, which is at 1,512’ MSL, for a density altitude of 4,089’ MSL. The discussion centered on the fact that the minimum requirements in the regs are pretty darned minimal.

    I ran a check of the airplane at its current stage, and the airplane actually climbs at about 760 fpm in the specified high-density-altitude condition, quite a bit better than a Cessna C-152 in the same conditions. However, there are a lot of simplifications and assumptions in the first-order sizing process, and one of the big ones is propeller efficiency, which directly impacts things like takeoff distance and climb rate. It’s completely likely that my design won’t quite climb like the first-order sizing says it will, once I’ve gone through the more-detailed second sizing and optimization. I can bring back some of that performance with changes to span, aspect ratio, and wing loading, but I need a specification that says, “good enough” to drive those calculations. I agree with the commenters that my current specification is inadequate, so I’m going to change it.

    While the Cessna C-152 is no rocket-ship, I’ve flown one enough, and on hot enough days, to know that its climb performance is “good enough” to be a worst-case-scenario for my little airplane. So I’m resetting my climb specification to be “matches or exceeds POH climb performance numbers for a full-gross C-152 under equal conditions.” Of course, on a cooler day, the airplane will climb much better. I have the 1978 Cessna POH for the C-152, so I'll be using the chart in there to get numbers against which to check my design.

    Range
    In one of the trade-studies I just ran, I looked at how the design range of the airplane affected the choice of engine. Within the ranges I tested, the result was, “not much”. My baseline engine is adequate to maintain specification performance all the way out to 500 statute miles. The question then becomes, is there anything I can actually do with some excess range, or would I just be adding it for the sake of adding it? While there’s a school of thought that says, “get as much range as you can,” you have to remember that this airplane really isn’t intended for lots of powered cross-country flight. That’s a secondary goal, behind low cost, short-range airshow trips, and self-launched soaring flights.

    I pulled out the charts and looked farther out than the distance to my current design-range airport, Fresno-Chandler (KFCH), to the north and east (To the west is nothing but Pacific Ocean, and to the south is Mexico). Is there anything farther out that would be good to do with this airplane?

    And the answer is, yes. I don’t know how I missed it, but there’s Bishop Airport (KBIH) just a ways north of Fresno-Chandler, and Mammoth Lakes (KMMH) just to the north of that. Bishop is a Mecca of soaring in the western United States - one of the best soaring spots on the continent. I must’ve been in “power plane mode” when I was thinking about destinations for my specifications list. Bishop would be someplace I’d simply love to go soar if I could get there easily. Mammoth Lakes is an area where I love to camp and backpack, and ski in the wintertime. I don’t think I’d be taking this airplane there a lot, but as it’s not far from Bishop, it’d be frustrating not to make the short jaunt over there if I’m in the area.

    I went back to Skyvector.com and plotted an “eyeball” flight from Corona Airport (KAJO) to Mammoth Lakes (KMMH), avoiding mountains, big airspace, restricted areas, and MOA’s. The result came out at just shy of 280 nm, which I’m going to call 300 nm, to which I’ll add the same 50 nm “reserve” I used before, for a total unrefueled range of 350 nm. That works out to just about 400 sm, which is the new total range requirement for my airplane.

    I should make note of the fact that both of these airports are quite high-altitude, with Bishop at 4,124’ MSL, and Mammoth Lakes at 7,135’ MSL. Wow. Both are noted for density-altitude issues, and I don’t see a reasonable situation where I can afford the kind of airplane (meaning engine) that could operate from these airports with impunity any time of any day, any season. Like others with low-powered aircraft, I’m going to have to be careful and plan my flights out of these airports for the early-morning hours, when the air is nice and cool. I won’t be changing my takeoff and climb specifications for these airports, and I will have to run some performance numbers before I go to either of them the first time, to make sure I have adequate margin to be safe.

    Engine BSFC
    Just a minor change here. I’ve been using 0.43, but with a little more research, I find a bit more scatter in the data, and that my current number might be a tad optimistic for a real-world ½ VW engine. I’m bumping this up to 0.45.

    Re-Sizing the Airplane
    Now that I’ve made these revisions to the requirements, what does that do to the airplane? If you recall, I’ve tied my three sizing-related spreadsheets together, so running a complete first-order sizing is a very quick and easy procedure. Plugging the new numbers in, I get the following results:

    The maximum wing loading that can meet the new climb specification is well above 7.6 psf, so the wing loading is still determined by stall and takeoff conditions. No change here.

    The new sized weights and characteristics are:

    W[SUB]0[/SUB] = 746 lbs.

    W[SUB]e[/SUB] = 477 lbs. (I know, I know. Just go with it for now, please.)

    W[SUB]f[/SUB] = 59 lbs.

    Backing these weights out through the “known” ratios for the airplane, as I did before, gives:

    Wing area = 98 square feet

    Span = 42 feet

    Fuel volume = 9.8 gallons (including unusable)

    Power required = 30 hp

    Not much change, really, and the baseline engine is still quite adequate. Yes, I’m still using the “old planes” empty-weight data, and not accounting for more-modern construction materials. We had a very lively debate about empty weight in the discussion thread, and I agree that I can do better than the empty weight number above, even using hand-layup composites. But I’m going to wait on revisiting that until after I’ve decided what materials I want to use to build this airplane and decide what it’s actually going to look like. Right now, I know I have a comfortable margin in empty weight between what meets the specification (the numbers above) and what’s possible for a good design for this airplane. I’d rather have margin than trust that “everything will come out just like I planned,” so I’m sticking with this empty weight number for now. You can trust that I’ll be looking at it a lot more closely once I’m into the more-detailed second sizing and optimization.

    Next Post(s): Listing out everything I know about the airplane at this point. Deciding some key things like landing gear arrangement, construction materials and methods, etc. Looking at some candidate overall configurations. Need all this to start drawing!
     
    Last edited: Sep 9, 2015
  3. Oct 2, 2015 #43

    Topaz

    Topaz

    Topaz

    Super Moderator Staff Member

    Joined:
    Jul 30, 2005
    Messages:
    13,727
    Likes Received:
    5,407
    Location:
    Orange County, California
    Interlude: Work happens.

    Sorry for the extended delay on this project. I have three major projects going in my business at the moment - a pair of catalogs, a series of seven books for a publishing company, and the set-up for another catalog workflow-automation project - and time has been at a premium lately. I've been doing some work on this airplane project, but it has been sporadic.

    I'll be back to it soon. The pair of catalogs finish up this week, and the other two projects by the end of the month.
     
  4. Oct 29, 2015 #44

    Topaz

    Topaz

    Topaz

    Super Moderator Staff Member

    Joined:
    Jul 30, 2005
    Messages:
    13,727
    Likes Received:
    5,407
    Location:
    Orange County, California
    Detailed Description of the Aircraft

    Before I start drawing, I want to bring together everything I know so far about this airplane, so that I have it all at my fingertips to guide my visual design work. Following that, some discussion about things that arise from the specifications and “knowns”. Knowing so much about the airplane before I start drawing is rather the point of the entire exercise. I’ll start off with the generalities from the beginning of the work, and then dive into the detailed numbers.

    Revised Requirements and Specifications
    I've done a lot of adjusting and changing to the requirements of the aircraft, as I learn more about it during the early design process. It's time to bring all those changes back together into the requirements and specifications document, which I have now done.

    As before, I'm not going to bother trying to arm-wrestle with the forum's table-creation tool for this. Instead, I've created a PDF version which, if you like, you can download from this link: https://app.box.com/s/h9cvfmu36r3rlkiq66n4zzynrhe44g83

    Technical Knowns

    Engine

    I wrote Scott Casler at Hummel Engines and requested more details about the 32hp 1/2 VW I've baselined for the project. Scott was very helpful and forthcoming, and I've learned some new things about my choice. Scott also provided me with a set of overall assembly drawings showing the major dimensions of the engine, which will be extremely useful when doing the initial layouts. All the numerical data is tabulated below:
    .
    • Type/Vendor: Hummel Engines 1/2 VW, two-cylinder opposed.
    • Takeoff Power: 32 hp
    • Continuous Power: 26-28 hp
    • Weight: 102 lbs.*
    • BSFC: 0.45**
    • Cost: $3,550 plus tax***
    .
    * This is the all-up weight for the engine with the options I selected (see below), minus prop and mount, as quoted to me by Scott.

    ** Scott doesn't have a specific BSFC number for the motor, but quoted me a consumption figure of "about 2 gph" at full power, and "about 1.6 gph at cruise." For design purposes, BSFC is usually given for full power (WOT), so I calculated a BSFC of 0.38 using 32 hp and 2 gph consumption. I've heard numbers like this for small VW installations before but, frankly, I'm a little skeptical that a real-world engine in a real-world installation in a real-world cowl can reliably attain that number after pulling a little airplane through the sky for a couple of years or so. I'm sure Scott is faithfully reporting actual numbers to me, and I'm not casting an aspersions on him at all. Since range is involved - a critical performance parameter - I'm simply going to keep using my older, more-conservative number for BSFC. The worst thing that happens there is that I have more range. What a crime.

    *** In my discussion with Scott, I let him know the kind of airplane I'm designing (light motorglider) and that I was interested in the starter-only option for the engine. Scott let me know that the starter in that option package is mounted on the front of the engine (prop end) and that it "does not allow for a clean cowling." Fudge. His recommendation was that, since I require an air-start capability, I instead select the starter/alternator option, which mounts those components on the back of the engine near the alternator. This option is about $300 more, which raises the total cost of the engine from $3,225 to $3,550. Still, I greatly appreciate him letting me know this up-front. It would've been maddening to receive the engine, discover the problem and ask him about it, only to hear, "Why, you didn't ask." I've known (ex)vendors like that. I'm glad Scott isn't one of them.​
    .
    Weights
    W[SUB]0[/SUB]: 746 lbs
    W[SUB]payload[/SUB]: 210 lbs (Pilot + baggage)
    W[SUB]empty[/SUB]: 477 lbs This is an upper bound, not the actual empty weight.
    W[SUB]fuel[/SUB]: 59 lbs (9.8 gallons) This includes unusable fuel, at 6% of total volume.​
    .
    Wing
    Wing Area: 98 square feet.
    Wing Span: 42 feet
    Aspect Ratio: 17.6
    Wing CL[SUB]max[/SUB] (flaps up): 1.10
    Wing CL[SUB]max[/SUB] (flaps down): 1.45
    Flaps: Plain, full-span flaperons or drooping ailerons and balance of span flapped.​
    .
    Next Post: List of specifications and knowns that have a major impact on the design.
     
    Last edited: Aug 30, 2018
  5. Oct 30, 2015 #45

    Topaz

    Topaz

    Topaz

    Super Moderator Staff Member

    Joined:
    Jul 30, 2005
    Messages:
    13,727
    Likes Received:
    5,407
    Location:
    Orange County, California
    List of Specifications and Knowns that Have a Major Impact on the Design
    This is a technique from Roskam's Airplane Design, Part 1. Basically, it's another "Stop and think!" exercise. I'm taking a look at what I know about the airplane and requirements so far and thinking through the ramifications of things that are going to have a big influence on the design. Some good insights come out of this process.

    Project Cost. I've already locked-in about $3,600 of the project's threshold $8,500 build cost, just with the engine alone. The rest of the airplane should be as small as possible (wings, etc.) and of relatively inexpensive materials. Anything that can be done at the configuration level to keep the cost down is welcome.

    Quick-build. The whole point of this airplane is as a quick and inexpensive learning exercise. Complicated construction techniques that require a lot of time-consuming effort are at odds with that goal, even if they offer other advantages. Within the boundaries of the threshold specifications, I'm willing to trade performance for fast-building and low-cost.

    Construction and storage space. The airplane will be built and stored in a single-car garage, which must also contain a single car (Porsche 914, so a small one). Here's the overview of the space again, with the car to scale.

    Storage Space.jpg

    This leads to some constraints on the size of the airplane:
    • No part of the airplane can be more than 19 feet long, when in storage configuration. Most should be 17 feet long or less when being constructed. For example, in the case of the fuselage, that means that it’s okay to be 19 feet long when built, including engine, spinner, rudder, etc., provided it’s 17 feet or shorter without those items.
    • Wings can clearly store flat up against one wall. I've shown that in green in the overview above. I've put them away from the entrance of the car so that they don't start to accumulate "garage rash". Keeping my wings smooth and clean is the key to keeping good soaring performance!
    • The space available between the top of the Porsche and the ceiling in the garage is about four feet. If the airplane would fit beside the car, it opens up that height to about five feet (I still have to be able to swing the car door open under the airplane). I've roughed out this volume in blue in the overview above. It should be noted that there is additional height available at the front or the back of the car. About another eighteen inches, when centered over the car. Clearly this is going to be tight fit, and juggling actual objects in the real space may prove to show a better way, but this is a good starting point and helps bound some airplane dimensions, depending on its shape.
      914-dimensions-copyright-porsche-downloaded-from-stuttcars_com-01.jpg
    • The garage door is eight feet wide. The airplane must be able to pass through this opening with the wings removed. This is also the maximum legal road width in California without a special permit, so these two requirements work together.

    Visibility from the cockpit. As a sailplane, this airplane will frequently be used in close proximity to the ground (ridge soaring) and other gliders (sharing a thermal, ridge soaring), so very good visibility, particularly forward and to the side, is necessary. Being able to see above the aircraft and, to the best extent possible, below it, is important as well. Any obstructions at all must be minimal, and a "necessary evil".

    Soaring performance requirements. My soaring requirements mean that drag reduction is a high priority in the design. Laminar flow airfoils for the flying surfaces, attempting to get some laminar flow on the fuselage, and minimizing flow separation at the wing-fuselage junction should all receive attention. Tempering this, I don't need competition class low drag. The SGS 1-34 is an example of the level of drag reduction for which I'm shooting. One consideration for soaring is cleaning up the engine cooling system when the engine is not running. The drag of air flowing through the cooling system may be unacceptable for soaring, and so the system may need some way to close off the intakes and possibly the exhausts of the cooling system. This should be kept as simple as possible.

    Cruise speed under power. Because of the drag reduction imposed by the soaring requirements, overall performance under power should also be okay. But there are some things to look at and consider:
    • Airfoil performance at design cruise speed. Airfoil selection for soaring will probably drive the laminar-bucket Cl into a fairly high number. For cruise, the airfoil should still be operating in the laminar bucket, but the airfoil will be operating at a much lower Cl than when soaring. I see two possibilities: 1) Select an airfoil with an extremely broad laminar bucket in terms of Cl, encompassing both design points. This may be very challenging. 2) Use my full-span aileron/flaps (or flaperons) as camber-changing flaps for cruise, moving the laminar bucket into the cruise Cl range at higher speeds. While not the ideal solution, it may be the more practical one. I'll know more once I get into airfoil selection.
    • Fuselage pitch angle at cruise. In a similar manner, the fuselage pitch angle (wing incidence angle) will be chosen to minimize drag at one of the soaring condtions. That will probably be L/D[SUB]max[/SUB], since it occurs at a higher speed than minimum-sink, and so parasite drag is more important. At powered design cruise, the fuselage would be flying more nose-down than ideal, generating more drag. The best way to fix this is the camber-changing flap scheme I already mentioned above. That's probably the way I'm going to go, since it helps both issues. The question to be answered is if the low-speed regime should have flaps down, or if the design powered cruise should have flaps reflexed? Something to look at in the next stage. Note: This is very much a sailplane/motorglider issue, because such aircraft have to be optimized for two design points (thermalling and cruising). For normal sportplanes, you're generally trying to optimize the airplane for the design cruise speed, and that's a single point which you'll use for the cruise portion of the airfoil selection process. If you decide you need camber-changing flaps for a sportplane for this kind of purpose, you're probably doing something wrong.

    Runways and possible foreign-object damage to the propeller. Looking at the condition of the runways I've selected (I collected this data earlier, and it's detailed there), it's clear that, while they're all paved, none of them could be called "great" in terms of being fresh concrete or asphalt, in extremely well-maintained condition. Most are listed as asphalt, in "good" to "fair" condition. Dirt, small bits of rock or asphalt, and other debris are likely to be present on many of these runways. This is common for the runways of most soaring operations, in my experience. Propellers are expensive. While still taking all the other configuration needs into account, I need to give consideration to protecting the prop for any configuration I choose, especially for pushers, where runway debris might be flung up from the landing gear into the prop arc. Ideally, this issue is resolved by the configuration of the airplane, and not by any special "shields" or protectors added to the design.

    Next Post: General design decisions, based on the discussion above.
     
    Last edited: Oct 30, 2015
    Himat likes this.
  6. Oct 30, 2015 #46

    Topaz

    Topaz

    Topaz

    Super Moderator Staff Member

    Joined:
    Jul 30, 2005
    Messages:
    13,727
    Likes Received:
    5,407
    Location:
    Orange County, California
    General Design Decisions, Based Upon the Discussion Above
    These are some items I want to determine in the airplane, which will help drive the ultimate configuration choice.

    Seating Position
    I’m selecting a supine seating position, typical of sailplanes, for drag reduction and pilot comfort. The seat back angle will be chosen by visibility requirements.

    Landing Gear
    I’m selecting mono-wheel main landing gear. I prefer this type of gear from my glider experience, and I find it very easy to use, especially in a cross-wind. Other benefits include low weight and lower cost (only one of each component). I would prefer that the main gear be retractable. I need the experience for my larger two-seat design. However, nothing fancy. No hydraulics, no electrics. Manual retraction only.

    The outriggers will be manually plugged-in as-needed, on the ground. I see retractable outriggers as a needless complication for this aircraft. For soaring flights, I’ll either be out with the soaring club and very likely have a wing-runner, or be operating in a sailplane environment (manual "taxi out") from a smoothly-paved runway where a wing-down takeoff should be a snap. Tip skids or small fixed tip wheels will protect the wingtip. For powered missions to and from "regular" airports, the outriggers can be plugged in to facilitate taxiing, and remain in-place for the duration of the flight. In powered flight, the drag penalty would be minimal. The vertical placement of the wing has bearing here - a high wing would mean longer, heavier outriggers. Whether or not that's actually an issue is something to address later.

    A fixed (non-retractable) steerable tailwheel rounds out the landing gear. Drag reduction should be a concern with this, and something that I will look at.

    Canopy
    Canopy choice is a strong cost driver in most situations. Normally, the budget constraints of the project would preclude a molded canopy, but I have an “out” in this particular area. My friend from Boeing and I committed last year to build a canopy molding machine that could be used for all of our aircraft projects. By the terms I set up at the beginning of this project, multi-use tooling doesn’t count towards the project cost total. Obviously the plastic used in the canopy itself will count.

    Wing Disassembly/Folding for Trailerability
    If you’ll recall, the pending closure of Skylark Field airport has made me put a much larger emphasis upon trailerability in this project. There aren’t any other affordable hangar or tie-down options in my area.

    This means the airplane’s wings will need to come off for trailering, and do so easily and relatively quickly. While I’ll leave details of how this will be done for later, some constraints are apparent right now. For example, I don't want to try wing folding. Too complex for this design, in my opinion. I’ve already determined that the wing span will be 42 feet. If I were to break the wings into two panels, each panel would be roughly 21 feet long, give or take a foot depending on the exact details of how they attach to the fuselage. Looking back to the available building/storage space in my garage, listed earlier in the project, the maximum allowable part length is 19 feet, with 17 feet being much easier to handle and store. Clearly, I will need to make three-panel wings for them to fit into the garage for storage - and to be built in the first place.

    Looking at another factor, maximum legal road width in California (and most of the country) is 8 feet. And the opening of my garage door is exactly 8 feet wide. If I decide to keep the middle panel of the three-panel wing permanently attached to the fuselage, its maximum span would be 8 feet, less a little for clearance that I’ll ignore at the moment. With that dimension fixed, how long would the outer panels be?

    Span = 42 feet

    Center panel length = 8 feet

    Outer panel length = (42-8)/2 = 17 feet

    Neat! Sometimes it just works out. Now, a “real world” outer wing panel will have some kind of attachment to pass bending and torsion loads into the inner panel. I prefer sailplane-like attachments, where a “stub spar” extends into the inner panel and is engaged with pins. That would lengthen the outer panel’s overall dimension, but it shouldn’t amount to two extra feet per panel. Total panel length for each of the outer panels should be somewhere close to, but less-than, 19 feet. That fits within my storage requirement. It slightly exceeds what I'd like for building, but the wings shouldn't take as long to build as the fuselage, and so are a little less-critical here.

    Next Post: Selecting primary construction materials and methods.
     
    Last edited: Oct 30, 2015
  7. Oct 31, 2015 #47

    Topaz

    Topaz

    Topaz

    Super Moderator Staff Member

    Joined:
    Jul 30, 2005
    Messages:
    13,727
    Likes Received:
    5,407
    Location:
    Orange County, California
    Select Primary Construction Materials and Methods

    First off, a couple of points of business.
    .
    1. I goofed. :emb: I should’ve had this piece of work done before I started the detailed description of the airplane and, having it here, after that process has begun, needlessly muddies that description. Things have been been beyond hectic in my “real world” life lately, and I’ve been less organized than I should be. This work needs to be included, and I can’t add a post earlier in the thread because of the way the forum works, so here is where it goes. Sorry!
    2. Some of this material, particularly with regard to molded composites, is hugely controversial here on HBA. My description here is my own interpretation of the issues, from my perspective, for this airplane project. I appreciate that it’s entirely likely that a “professional” scenario, in a proper aircraft factory or university lab, would be an entirely different case. My take is from the perspective of one completely average guy building a one-off, single airplane for himself, for the first time, with very significant time and cost constraints. Your mileage may vary, of course, and I know for a fact that, in the case of several members here, it most certainly does. We’ve discussed that at length here. I know your opinions and, while I understand and respect them, I disagree and there is no need to restate them in the discussion thread for my project. While I normally enjoy input in the discussion thread very much, let’s not re-ignite the “"molded/moldless is cheaper/quicker” debate there. If you just can’t stop from correcting me from the vast litany of errors you think I’ve committed below, please start up a new thread on the subject and the world can decide for themselves. I have no desire to have that debate yet again.

    I now return you to our regularly scheduled thread.

    While a basic “airplane” shape can obviously be built out of any of the common construction methods, the specific material(s) and method(s) will have an effect on the final detailed shape of the aircraft. For example, barring expensive draw-forming methods or a high-skill tool like an English Wheel, building complex compound curves in sheet metal is pretty much impossible.

    In my actual design workbook, I have an outline-style listing of each material considered, and the pros and cons of that material in the context of my design. I personally find that kind of presentation great for examining alternatives and making a decision, but it’s a bear to code it in here on the HBA forum and it doesn’t provide context, since the context is all in my head. But I think that you can’t justify a good decision unless you can explain it, so I’ve retyped that material in a more forum-friendly format, and added the background context. You’re getting a much-expanded form of the conversation I had with myself on this topic. I’ll warn you right now - this is a long post.

    Tube and Fabric
    This is the traditional Cub-style build, with a 4130 steel-tube welded frame, wood wings, and a fabric covering. It’s very light, and I do have some woodworking experience for the wooden parts. However, the aerodynamics possible with this method are wholly inadequate for the kind of soaring performance I want, barring a hugely complicated build with hundreds of small formers and stringers to push the fabric covering into the right shape. Even so, the level of smoothness possible would never allow any laminar flow. While I have started to learn welding, I don’t consider myself competent to weld an entire structural tube frame. I’d likely have to farm that work out. The last consideration is that cutting the wood parts is going to be noisy saw and router work. Recall that keeping the build noise down is a requirement of my project.

    All Wood
    This type of build uses wood frames and a few longerons and/or stringers, and then covers most of the aircraft in varying thicknesses of plywood. Aerodynamically, this can be surprisingly good if care is taken during construction - even competition sailplanes were once built this way. It’s much harder to build a low-drag fuselage this way, because of the compound curves. That means steaming and bending - a skill I have not yet acquired. But I do have some general woodworking experience, so most of the build would be familiar territory for me.

    All-wood builds are generally either light and very complex (a lot of little parts) or heavy and simple. Rarely are they both light and simple, if the aerodynamics are any good. And again, we have the noise problem of cutting out the parts with saw or router, and finishing all the edges and surfaces.

    Riveted Aluminum Sheet
    Traditional riveted aluminum sheet construction can be one of the easiest, lightest, faster-building methods out there. One has only to look at Van’s RV-x series to see that. There’s a lot to be said for this method, and the fact that it’s been the dominant airplane construction technology for decades bears that out. If I were doing a small pure-sport airplane like a Hummelbird, this would probably be the method I’d choose. Obviously the designer of that airplane feels the same way.

    The problem in my application here is aerodynamic. The style of construction for an RV-x is unable to provide the low-drag, laminar-flow flight surfaces and low-drag fuselage surfaces that I need for this project. Sheet metal construction can be brought to that level of accuracy and smoothness, but it requires relatively thick, stretch-formed skins, especially for the wing and tails forward of the spar. Only one homebuilt design of which I’m aware used this technology, and that’s the SX-300 by the now-defunct Swearington kitplane company. The tooling requires extremely precise solid metal forms and industrial-grade stretch-forming equipment to stretch the final skins over the forms. Stretch forming does allow compound curves to be developed, which is a plus.

    The biggest “plus” of sheet-metal construction is that it’s going to be one of the two lightest construction methods of all the ones I list. It’s also one of the easier building methods, especially with matched-hole techniques. Structural “pulled” rivets are also relatively quiet to set, as opposed to “set” rivets, and cutting thin sheet can be done with shears and cutters, rather than saws. Pulled rivets on the exterior of the aircraft are a big source of drag - even if countersunk - because of the roughness created by the stub of the mandrel. Unfortunately, I don’t have a lot of sheet-metal tools or experience. This would be a new method for me.

    Moldless Full-Depth Foam and Fiberglass
    This is the method pioneered on the first composite sailplanes by the German akafliegs and made hugely popular in the 1970’s and 1980’s by Burt Rutan’s EZ series of canard homebuilts. The Quickie Aircraft Quickie 1 and Viking Dragonfly are two other kitplanes that were entirely built with this technique. Numerous others have used it for wings and flying surfaces.

    The big plus of moldless composite work is that it’s probably the fastest way to build an aerodynamically smooth airplane with lots of compound curves. Rutan used this method at Scaled Composites for early prototyping work, simply becuase it’s one of the fastest build methods period, requiring virtually no tooling or jigs. For me, there are two added bonuses: Building an airplane with this method is virtually silent, and it’s a technology with which I have some dedicated aircraft-building (models) experience. I’m very comfortable with it.

    This method is heavier than both riveted sheet metal and tube-and-fabric, but probably lighter than all but the most complex all-wood structures. At larger sizes (bigger than most homebuilts) the full-depth foam core starts to become a weight issue, but at the size of my aircraft, this method is actually lighter than all but the most-optimized molded sandwich composite structures. The square-cube law is my friend in this case.

    Let’s take a moment to talk about finishing. “Everyone knows” that moldless composite airplanes are a nightmare of sanding and filling and sanding and filling and sanding and filling… I’m going to say this once here, although I’ve said it several times elsewhere on HBA: What “everyone knows” about finishing moldless composite airplanes is completely wrong. If one takes care during the entire build process to produce smooth foam cores and fiberglass laminates without wrinkles and with proper transitions between laminates, “finishing” is limited to simply filling in whatever weave texture is left on the outer layer of laminate and sanding that layer of filler smooth. A couple of days for a wing, maybe four - not weeks or months. Then it's primer and paint like anything else. If you’re having to spend months sanding and filling your moldless fiberglass airplane, you’ve done the earlier steps horribly wrong.

    Moldless Full-Depth Foam and Carbon Fiber
    Same as fiberglass construction above, except carbon is used as the fiber reinforcement. The addition of relatively inexpensive carbon fiber fabrics to our available materials is huge. Carbon is awesome not because it’s strong (although it is) but rather because it is stiff. A structure can deflect less under the same load with far less material, and therefore weight.

    Carbon fabrics can be used for moldless laminates, just like fiberglass. You get all the good things of moldless composites, with all the extra benefits of carbon. Structures are stiffer, and lighter for the same strength. The fabrics themselves are more expensive than fiberglass, but you use less. Most comparisons of which I’ve heard report that the cost difference ends up being a wash.

    So why didn’t the moldless composite world switch over to carbon, en mass? Two words: Quality assurance. Not of the fabrics themselves, which are probably as good or better than the glass fabrics available, but quality control for the individual airplane builder. Fiberglass, when “wetted out” by epoxy, turns nearly transparent. You can see every bubble, fiber or dirt inclusion, areas that are “dry” (inadequate amount of epoxy), or delamination. Inspection is easy, entirely visual, and quick. Carbon fabrics, however, remain as opaque when wetted out as they were dry. You have to be much more careful at the laminating stage to make sure you’ve gotten enough epoxy into the fabric, haven’t included any bubbles, and so on. It’s not at all impossible, but it takes practice and experience. In a regular factory situation it’s a non-issue. They use fabrics pre-impregnated with exactly the right amount of expoxy and sophisticated methods like ultrasound for quality assurance of their parts. For a first-time homebuilder, it’s a little more problematic.

    Except for the quality-assurance issue and the much better stiffness that carbon provides, this method/material combination has all the same plusses and issues as moldless fiberglass construction.

    Molded Sandwich Foam and Glass/Carbon
    This is modern sailplane construction, with foam core/carbon-fiber sandwich shells molded to shape in large female molds. When taking aerodynamic and structural qualities into consideration, it’s simply the best, particularly with carbon fiber as the laminate.

    The results, in carbon, are as light or lighter than any form of metal construction. Even the fiberglass version is nearly as light as the most-sophisticated sheet metal build. If the plug-and-mold process of building the molds is done carefully, the aerodynamics are second to none. For very small airplanes, however, an interesting thing happens. Once you factor in the weight of inside skins, closures, and so on, (all of which are heavier than foam) sandwich construction can actually be heavier than full-depth foam. It’s the square-cube law at work again. I don’t know where the dividing line really is, other than “somewhere at the high-end of Glasair/RV-x size territory”, and it will vary from design-to-design,

    But there are downsides. It all comes down to the molds. In the case of fuselage molds, you’re essentially building the airplane twice, or possibly even three times. In traditional molding workflows, you build a “plug” - a thing that is exactly the same shape, size, and smoothness as the airplane part you want to build. Then you laminate a female mold onto this plug. Pull the plug from the mold, add a bunch of really stiff structure to keep the female mold in shape as you work, and then you laminate up the airplane skin and structure inside this female mold. It’s not actually as simple as I’ve just listed it out - you have to sand and fill the plug to get it to the right shape and smoothness (about the same level of work as on a moldless aircraft, done well), and usually the female mold needs some finish sanding and “dressing” to get it ready to go. You have to repeat this process for every molded part on the entire airplane.

    Modern molded airplane technology has made some advances on this, generally eliminating the plug step by carving the female mold directly with CNC equipment. You then hand-laminate the actual mold surface as before. However, this time you’re not laminating up against a smooth male plug, you’re laminating on a negative of that shape, and the actual mold surface will need much more hand finishing and dressing than one done against a finished plug. There’s no way to avoid that labor, even with a CNC mold. You just transfer it from the plug to the mold itself.

    With parts like wings and tails, there are easier methods that do not require a plug, and only a minimal mold. Most of the issues above arise solely with fuselage parts.

    All of this takes time. It takes material. And those mean money. If you’re doing it the “modern” way, you have to have access to a six-axis CNC foam cutter mill large enough to encompass the entire part - which can be quite large if you’re talking about an entire cowl, and worse if you’re doing a fuselage half. It has to be accurate enough to hold laminar-flow tolerances. Plans and even kits for hobbyist “maker” CNC foam mills exist on the Internet for machines large enough for this purpose. Which, of course, you would have to buy materials for and build itself. Or you can contract the CNC carving out to a commercial firm, and pay them for the service.

    It is my considered opinion that all of this, taken in its entirety, is going to take longer and be considerably more expensive than Rutan-style moldless construction for the solo one-off airplane builder. In the case of flying surfaces, the total trade-offs may be slightly in the other direction, since virtually all sanding and filling is eliminated for those parts. It is my opinion that, over the entire airplane, the total balance of molded versus moldless favors the latter in terms of construction time and expense.

    Final Construction Method/Material Selection
    It’s not going to be any surprise to anyone that I’m selecting moldless full-depth composite as the overall construction method. For me, as that “solo average Joe building a one-off airplane for himself,” I don’t want to spend the time, materials, and money on plugs and molds, garage-sized CNC machines, or professional CNC services to carve those molds. Not for an $8,500 airplane. I want to stay clear of that time and expense even if it means I have to spend a few extra days or even weeks filling and sanding by hand. I personally don’t see the extra time, money, and material in molds as a valid trade-off for a little elbow grease on an airplane like this. The final consideration in the decision is that I’m already familiar with moldless and am comfortable with it. That’s extremely important for any builder.

    I’ve disqualified tube and fabric and riveted sheet metal as being aerodynamically inadequate for this particular application. Wood (and the wooden portions of tube-and-fabric) is simply too noisy, for the many parts that go into an airplane. The reasons for my concern about noise were detailed earlier.

    Now for material selection. From a purely structural and engineering point of view, carbon fabric is the clear winner. It has all the good qualities of fiberglass, only better. However, strutural and engineering considerations are not the only valid ones. Cost is pretty much a wash between carbon and glass, from everything I’ve read, so that’s of no help. That leaves production issues. I have to be able to build this thing, with my own two hands, in my garage, and be able to make sure that I’ve done an adequate job before risking my life flying it. I’ve italicized that last sentence because it’s what drove my final decision to use primarily fiberglass in my airplane. I’ve outlined the quality-assurance issues with carbon versus glass above, and I do not have what I feel is adequate experience with laminating carbon fabrics to feel really solid assurance that I’ve done every laminate properly, over a large area, with no voids, air bubbles, or delaminations.

    With the long sailplane wings on this project, it may well be that fiberglass is not adequate for high-stiffness parts like the main wing spar. Use of carbon fabric or pultruded carbon rods for such parts is not out of the question. For relatively small-area parts like spars, I can probably develop enough skill with that specialized kind of layup. If not, pre-made pultruded carbon rods will do the trick. That’s something to be determined during the later “preliminary” design phase between “Conceptual” and “Detailed”, and I will not be touching that subject during this conceptual design phase.
     
    Last edited: Oct 31, 2015
    PTAirco, Himat and RPM314 like this.
  8. Nov 1, 2015 #48

    Topaz

    Topaz

    Topaz

    Super Moderator Staff Member

    Joined:
    Jul 30, 2005
    Messages:
    13,727
    Likes Received:
    5,407
    Location:
    Orange County, California
    Select the Gross Configuration
    I’m selecting a conventional one-main wing, one aft tail configuration for this airplane. Here's why.

    The list of possible configurations.
    1. Conventional
    2. Flying wing/Tailless
    3. Canard/Tandem/Joined Wing/Biplane/Three-surface
    4. Lifting body

    Lifting bodies: While I'm fascinated with designs like the Facetmobile, I know from my earlier work that, to reach all the performance specifications for this design, I need the airplane to have an aspect ratio of about 17.6. A high-aspect-ratio “lifting body” is a contradiction in terms. Cross this one off the list.

    Canard/Tandem/Joined Wing/Biplane/Three-surface: For my purposes, these types are the same, in that the primary lift system is split across two or more flight surfaces. Lift theory shows that, for a given wing area, a monoplane will have lower induced drag than a multi-surface lift system, unless span is constrained. Low induced drag is critical for a sailplane or motorglider. Joining the lifting surfaces (“Box plane”, etc.) can reduced the induced drag for multi-surface designs, but not to the level of a monoplane of the same area and unconstrained span. Is span constrained on my design? Not within the limits imposed by my requirements. I need 42 feet of span to get the right aspect ratio with 98 square feet of area, and that amount of span will fit within my build and storage space.

    The sub-set of multi-wing aircraft where the wings are spaced fore-and-aft of the center of gravity (canard, tandem, joined wing, three-surface, etc.) can be designed so that they are stall-resistant. We’ve discussed this at length here on HBA, and while such stall-resistance might be beneficial for some kinds of aircraft, in a sailplane, full pitch control is needed right up to the stall, and that’s not possible with any form of aerodynamic stall-resistant design.

    For both of these reasons, I struck the entire class of multi-lifting surface designs.

    Flying Wing/Tailless: This was a more challenging choice. I like tailless aircraft for a variety of reasons. For one, there’s simply a lot less airplane to build, and that means less money spent - a big attraction on a project like this. The reduction in parasite drag possible with a flying wing would be a big help during powered cruising flight, enabling more speed for less power. The induced drag penalty inherent in tailless aircraft may or may not be an issue in this design, although I’m going to be using a fairly small engine, in the entire scheme of things, and I’ve already struggled a bit to keep the climb rate up and takeoff distance reasonable in the hot-and-high conditions I’ve specified.

    I’ve done enough research and design studies with the tailless configuration that I don’t accept the old saw that flying wings inherently have some issue with stability and control. The fact is, they fly by the same physics as a Piper Cub or any other airplane. They are however, much harder to design because so many design issues must be integrated together into the wing. With a low-power design such as mine, the difficult trade-off between factors promoting high efficiency from the wing and good stall and post-stall behavior make the task very complicated.

    In the end, I decided that, despite a big possible reduction in build cost, the design issues are more complex than I want to tackle with this project. I don’t think it’s impossible to do this airplane as a flying wing, but I want to keep this as simple as possible. Yeah, too late, I know. :gig:

    Next Post: The Inspiration Wall
     
    Himat, danmoser and RPM314 like this.
  9. Nov 6, 2015 #49

    Topaz

    Topaz

    Topaz

    Super Moderator Staff Member

    Joined:
    Jul 30, 2005
    Messages:
    13,727
    Likes Received:
    5,407
    Location:
    Orange County, California
    The Inspiration Wall
    I'm a graphic artist by trade, both design and production, and a tool I learned from the graphic design half of that profession is the “inspiration wall.” When you have a task at hand that demands synthesis of many factors, find as many images as you can that apply to the situation in some way. Put the pictures up on the wall and take time just looking at them. Take a lot of time. Doing so makes connections in your subconscious, and helps you see issues much more clearly than you could with simple visualization in your mind. In the case of airplane design, I chose pictures of examples of my candidate configurations that are already flying, pictures of some aspect of a design that I find interesting in light of my design goals, possible solutions to configuration issues I already know about, or simply an airplane that reminds me of my design goals in some way. I have a big magnet board behind my desk in my office, and I use that for inspiration walls for work projects and personal projects like this one.

    I’ve put together a “virtual” copy of my inspiration wall, which you can view here: https://drive.google.com/folderview?id=0B6fS_b7TW9U2NjdkenMtdU5mdUU&usp=sharing

    The images aren’t in any particular order, other than alphabetically by file name. I’m hosting my wall on Google Drive because I don’t want to use up this much of Jake’s storage. The link should allow you to view the Google Drive folder that I'm using as an inspiration wall. If it doesn't work, let me know.

    Photo credits
    Most of the images were taken from the Internet (my usage here falls under the Fair Use clause of the US Copyright Act), but there are a number of images that were either posted on HBA by members or sent to me by members privately by e-mail. I asked permission of the two members involved before I made this post today, and received permission to use their pictures in this way. I want to give appropriate credit for the images, and I thank you both for the inspiration and for allowing me to show your work to everyone!

    The E-ticket is a creation of our own Victor Bravo. There are several really clever things going on in this one view.

    The images of the Macro airplane are from its designer, our very own “Head in the Clouds”, who is showing his current build in the “DooMaw” build log thread. Those of us participating in the “VP-21” thread could do a lot worse than looking at this little airplane for inspiration. In my case here, in the course of my various design studies I’ve wrestled mightily with “scope creep”. The airplane always gets bigger, fancier, higher-performing, more-complex, and more-costly. The Macro is a wonderful example of the benefits of keeping it light, keeping it simple, and keeping it fun!

    The DS27 perspective view is one of my own earlier design studies. Clearly based on the work of Dipl. Ing Richard Vogt and the WWII-era Blohm und Voss design team, in particular the BV-141, which is also pictured. It's one way to have both a tractor propeller and a great view forward.

    If I've overlooked anyone whose images are included in this group, please contact me immediately so that I can give you the proper credit for your work.

    I’ve spent a lot of time looking over these images and using them to guide my final detailed configuration choice. They’ve been a lot of help in that regard. I don’t know that I can really give you a commentary on the entire collection, but I’ll do my best to answer any questions in the discussion thread. Any questions except for one: “Which of these did you pick for your final configuration choice???”

    The answer to that question is, yes, I’m still a bit ahead of what you’re seeing here, and I’ve narrowed the pack down to two final candidates. I’m strongly leaning towards one of them, but I haven’t made the final decision, and no, I’m not tellin’ yet! :gig:

    Other than that, you can read my two recent posts regarding design issues (#45 and #46) and review these images in that light. It’s exactly what I’ve been doing.

    Next Post: Define the major wing geometry, so I can start sketching candidate configurations.
     
    Last edited: Nov 7, 2015
  10. Nov 8, 2015 #50

    Topaz

    Topaz

    Topaz

    Super Moderator Staff Member

    Joined:
    Jul 30, 2005
    Messages:
    13,727
    Likes Received:
    5,407
    Location:
    Orange County, California
    Baseline Wing Geometry
    As I've reviewed my inspiration wall, I'm getting ready to start sketching out four candidate configurations so that I can see them in scale, and also so that I can show what I'm looking at here. Since I know a lot about the geometry of the wing at this point, it's a good place to start drawing. Pulling the known dimensions from my earlier analysis:

    Span = 42 feet

    Area = 98 square feet

    Aspect Ratio = 17.6

    While an elliptical wing is the "ideal" for low induced drag, they're devilishly difficult to build. Even competition sailplanes break the wing up into two or three straight-tapered panels, coming close to an elliptical shape while being easier to manufacture. Since top-notch performance isn't a requirement for my airplane, but moldless composites make taper relatively inconsequential, I'm going to choose a two-taper wing instead of three. A "rule of thumb" formula for minimum induced drag, in both Fundamentals of Sailplane Design by Thomas (p.99) and Sailplane Design by Pajno (p.123) goes as follows: Inner panel taper is 0.8, outer panel taper is 0.4 (from the root), and the "break" between the two panels occurs at 60% of the semi-span, from the center line of the airplane. Research has also shown that keeping a straight trailing edge and tapering the leading edge only can produce slightly lower induced drag than a pure-elliptical planform (Thomas, p.105, Pajno p.108). The same holds true for dihedral - a polyhedral approaching a elliptical curve when viewed from the front can also reduce induced drag. The former - planform - is easy to build, while polyhedral is more complex a task. I've decided to stick to the easy road and make the trailing edge straight and perpendicular to the centerline, but use a straight dihedral.

    A few dimensions had to be calculated from these choices. I built up a little spreadsheet to calculate the chords at the root, taper-change, and tip while holding the specified span and area, to guide my drawing. The root chord ends up being 3 feet. The spreadsheet is not really much of anything - you could easily do this with calculator and a pad of paper - so I'm not going to post it here. If you want a copy, PM me.

    There is a center section that remains attached to the rest of the airplane, and this will be 7.75' in span, to allow a little clearance for my 8' wide garage door and 8' maximum legal road width in California. The wing panels outboard of this section remove for trailering.

    Putting that all into a picture, here's my baseline wing. All my candidate configurations will use this same wing geometry. And yay! The first actual drawing for part of my airplane! :grin:

    DS54 Baseline Wing V1.jpg

    EDIT: Ugh. This image doesn't show up very well. The forum software is sizing it down. If you want to see a larger version, click here.

    All the control surfaces and drag brakes are completely notional in this drawing. Eyeballed in, and don't necessarily reflect the final size. While Thomas (p.127) suggests that sailplane ailerons are generally about 15% chord because of the extent of laminar flow over the wings, I've drawn these at 20% for now, just because I think 15%-chord ailerons look silly. I'll do the math later and redraw as necessary. I haven't decided if these will be ailerons and flaps, or nearly-full-span flaperons. I've drawn a split in them at the change in taper, to keep the hinges simple. Again, this is notional and not necessarily the final design. That 6 foot tall drafting dummy that shows up in airplane drawings has made his appearance, to give a sense of scale.

    No, I have not chosen an airfoil yet. That comes in the major analysis after I've selected a single final configuration.

    Next Post: Review inspiration wall and requirements, and choose no more than four candidate configurations. Sketch them and discuss.
     
    Last edited: Nov 8, 2015
    Himat likes this.
  11. Nov 11, 2015 #51

    Topaz

    Topaz

    Topaz

    Super Moderator Staff Member

    Joined:
    Jul 30, 2005
    Messages:
    13,727
    Likes Received:
    5,407
    Location:
    Orange County, California
    Interlude: A small adjustment

    Over in the discussion thread for my project, we got into a discussion regarding the span of the center panel of my baseline wing design, starting about here. I had set this at 7.5', so that there was some clearance passing the airplane through my 8'-wide garage door for trailering and storage. Some members here expressed concern that the clearance wasn't big enough for day-to-day operations. Various solutions were offered, ranging from simply shortening the center-section span, on up to making this a four-panel wing with the two major panels meeting at the fuselage and large-ish outer panels coming off so that the pieces would fit inside my garage.

    After some back and forth and some recalculating on my part, I decided that shortening the center-section span to 7' was the way to go. Overall span is set by my performance calculations at 42', and remains the same. The outer panels, with an 18" stub-spar continuing on into the center panel to join the wings, are now exactly 19' in length, which is the largest size that can be accommodated in a standard-sized US garage (20' deep).

    I'm not going to provide an updated drawing of the baseline wing here, but you'll see the change when I post drawings of the candidate configurations I've been exploring.
     
  12. Nov 28, 2015 #52

    Topaz

    Topaz

    Topaz

    Super Moderator Staff Member

    Joined:
    Jul 30, 2005
    Messages:
    13,727
    Likes Received:
    5,407
    Location:
    Orange County, California
    Select the Final Detailed Configuration

    Candidates
    After reviewing my inspiration wall, and considering the images I posted there in detail over a couple of weeks, I chose three candidate configurations to explore for this project. My goal now is to describe and refine these towards a single final choice, which will then be developed into finished form for this study.

    Please read this note first: The drawings of the three candidates below are not final “completed designs”. These are really fancy “back of an envelope sketches”, done “pretty” because I’m a graphic artist by trade and have a professional aversion to presenting something as rough as a pencil sketch in public. As such, they only show the concept of the particular design accurately, not the details. You already know how well the wing is actually refined in these drawings (they all share the same baseline wing). I have not done a weight and balance for these drawings - position of wings and such are by “eyeball”, and could very well be “off”. I did a very rough volume sizing for the horizontal tail for candidate 1 based on the drawing, just to get the tail size in the ballpark, and then reused that tail as-is for the others. (I’ll re-do that calculation in detail for the final winner, and I’ll explain it then.) Vertical tail sizes are by eyeball alone. The one concession to accuracy in these sketches is that they are to scale. I’ve found that hand-sketches and mental-visualization tend to be overly forgiving when trying to fit components into a design, and I always do even my “envelope sketches” to scale where possible. I recommend the same to everyone.

    All three concepts are 18 feet in length, use the same wing, and the same rough “envelope” sketch of the Hummel ½ VW engine, based upon the drawings provided by Scott Casler at Hummel Engines. The main gear tire is simply an RF-4 tire, since that’s a similar airplane in a similar weight class. It looks a little large relative to the airplane, but you have to remember that there is only one main tire, which has to support the entire aircraft. The pilot is scaled to 6 feet tall. The propeller is 54 inches in diameter, with a tail-up ground clearance of 5 inches in the first two candidates. In each of the drawings, the green rectangle is baggage, while the blue one is fuel.

    As rough conceptual sketches, all three could use some pretty thorough aesthetic refinement, although I have to say that Candidate 1 is pretty snappy-looking to me, even at this stage. Consider these to be “suggestive” of a potential final design derived from each, not a cast-in-stone final “look”.

    The side-views are to a larger scale than the plan views, because of limitations in the way the forum will allow me to scale the latter. The “DS Candidate” tag in each view is scaled proportionately, so it should give an indication of the size of the aircraft in each view. For those interested in software, these were hand-drawn in Adobe Illustrator v19.1.1.

    So here we go...

    Candidate 1 - Conventional

    DS54-C1-Side-Exterior.jpg
    DS54-C1-Plan-Exterior.jpg
    DS54-C1-Side-Interior.jpg
    DS54-C1-Plan-Interior.jpg

    I’ve always been attracted to unconventional designs, and I doubt I’m alone in that here on HBA. However, the tried-and-true conventional layout has some pretty strong things going for it, in that the body of design methodologies out there are pointed at this configuration, making it very low-risk. It’s hard to argue with the fact that, barring a pretty sophisticated structure and structural analysis, the conventional configuration can be the lightest option.

    Candidate 1 is essentially a smaller, more-sophisticated RF-4, SFW-31, or ASK-14. Mark Calder’s Robin is another take on the same idea. His emphasis was placed on fitting within FAR Part 103, whereas mine is placed on higher soaring performance, but they're essentially very similar aircraft.

    Plusses of this layout are its relative structural simplicity, probable light weight, simple control and throttle runs, and the short outriggers that the low wing allows. There is ample room and easy access for the baggage. The tractor propeller helps reduce or prevent foreign-object damage to the propeller on poor runway surfaces.

    The biggest of the weaknesses of this candidate is, to my mind, the relative lack of forward-and-down visibility. While I’ve provided an honest 10 degrees of downward visibility over the centerline of the nose, the cowling cheeks block the view down and to the sides a bit, and the wing blocks the view down and to the side. On the other hand, the view upwards from the centerline of the aircraft is completely unobstructed.

    Other relative weaknesses revolve primarily around drag, especially under soaring conditions. The forward engine pretty much rules out any laminar flow on the fuselage, and the low wing is draggier and should have small fillets to help with flow near the wing-fuselage junction. Yes, as-drawn, this may have some spin recovery issues because the horizontal tail might blanket the rudder in a stall, but that’s a matter for more-detailed design later.

    Those of you who insist on housing the fuel in the wings will be disappointed in this design - it uses a Cub-style forward fuselage tank. That doesn’t bother me at all, but I know some of you will be getting eye twitches over it.

    Candidate 2 - Asymmetric

    DS54-C2-Side-Exterior.jpg
    DS54-C2-Plan-Exterior.jpg
    DS54-C2-Side-Interior.jpg
    DS54-C2-Plan-Interior.jpg

    If I got doubtful looks for the fuel tank in Candidate 1, this one ought to have some of you questioning my sanity altogether! This concept is based on an earlier design study (DS27) I did many years ago for a two-seat tandem sportplane. I always thought that design would also make a nice motorglider. (I'd post an image directly here, but the forum software only allows twelve attachments per post, and I need those for my candidates!)

    That design was, itself, based on the WWII-era Blohm und Voss BV-141, a contender for a Luftwaffe reconnaissance aircraft requirement that was ultimately won by the Focke-Wulf FW-189 Uhu.

    The BV-141, designed by the gifted aerospace engineer Dipl-Ing. Richard Vogt and his team at Blohm und Voss, completely defies conventional thinking about airplanes. Although it may seem difficult to believe, the airplane actually flew very well. Instead of making the airplane “fly crooked”, the asymmetric design actually cancels P-factor and the airplane flies straight even during power changes! (I can expand on this in the discussion thread, if you like.) My interest in this configuration stems from the combination of a tractor engine installation - great for preventing foreign-object damage to the prop on rough fields - with an unobstructed forward view for the pilot. The cowling blocks the view somewhat to one side, but the critical forward and down sector has the same view as a conventional sailplane. I checked with our own SVSUSteve a while back, and fears of a thrown blade impacting the pilot pod have no particular basis in fact. GA airplanes almost never throw a blade, and there’s about a 60-in-365 chance that the blade will hit the pilot pod if it happens anyway.

    The pros of this candidate are visibility, visibility, visibility, while maintaining good FOD characteristics. This design also has ample room and easy access to the baggage. This is the only one of the three candidates that isolates the fuel (blue rectangle) from the pilot entirely, so it’s clearly the best if you’re into that.

    Downsides of this candidate are its relative structural complexity (compared to Candidate 1), probable higher weight, and the complex airflows in the area where the two “fuselages” intersect the wing. Control runs to the tails all have to take two 90° turns. The canopy is pretty massive, making the tooling for it pretty large, too.

    And then there’s the problem of the main landing gear. I have no idea how to make this retractable and keep the system light and simple. The only really viable way to retract the gear is sideways, into the engine fuselage between the engine and fuel tank. However, the doors and covers for doing that are pretty complex in any variation that I can conceive, completely inappropriate for this low-cost design. I strongly suspect that, for this project, I’d have to keep the main gear on this candidate fixed and faired. I don’t really like that.

    Candidate 3 - Pod and Boom

    DS54-C3-Side-Exterior.jpg
    DS54-C3-Plan-Exterior.jpg
    DS54-C3-Side-Interior.jpg
    DS54-C3-Plan-Interior.jpg

    I’m sure you noted all the images of the Stroknik S-2 and its derivatives in my inspiration wall. It’s hard to argue with success! The S-2 was the first homebuilt motorglider to complete flights winning all three FAI Diamond badges. That says a lot about good soaring capabilities.

    This is the only pusher in the list. It gets its FOD protection for the prop by virtue of the tail boom acting as a shield against material thrown up by the main wheel. In every other pusher configuration I looked at, FOD on the prop was a serious issue on poor-quality runways, requiring “fenders” or other shields on the main wheel, and still not providing really good protection.

    Plusses of this candidate, like Candidate 2, mostly revolve around visibility. As-drawn, this design provides about 13 degrees over-nose visibility, and completely unobstructed visibility down and to the sides. Visibility up and back are compromised by the wing, but that’s a less-important sector. The cowling acts like a bullet fairing for the rear wing-fuselage junction, and this design has the lowest wetted area of the three, as well as good chances of laminar flow on the front of the pod. As a result it could have the lowest overall drag of the three.

    This candidate has some challenges. It’s probably about the same weight as Candidate 2, which is to say, heavier than Candidate 1. While the cowling works as a bullet fairing, the airflow in the wing-fuselage junction is still pretty complex, with a lot of opportunity for flow separation. This has been a problem on many pod-and-boom designs, so it’s fairly high-risk. I don’t like the “L-shaped” firewall necessitated by positioning the fuel underneath the engine, and the baggage will need some kind of restraint to keep it from crashing forward into the pilot’s head in an impact. In fact, this is probably the “least crashworthy” of the candidates, because the thing doing the energy-absorption between the engine and the ground is the cockpit. I’m also not exactly sure where to put the fuel filler. Under the wing is awkward, and anyplace else is worse, from a fire-safety standpoint alone. When plugged into the wings for powered-cruise flights, the outriggers are going to be really long!

    Something came up in the layout of this concept that validates my use of scale drawings, and also ruled out all the other “pusher” engine installations at which I looked. To maintain reasonable prop efficiency, I placed the prop about 0.6 of the local wing chord behind the trailing edge. That’s a minimal separation - it ought to be about 0.7c. But pushing the engine aft starts pushing the CG aft - and that means a shorter tail arm and bigger tail for the same static stability. I just eyeballed that effect here, but this wing is about a foot aft of the wing position in the others, even so. My design is length-limited by the build and storage space to about 18 feet (all three candidate designs are exactly 18 feet long in these drawings), so I do not have the option to simply lengthen the tail. The Strojnik S-2 is nearly four feet longer than even the maximum length I can accommodate in my garage.

    So there they are! My job now is to look these over and decide which is the most promising for further development. Which would you choose, and why? I’m itching to satisfy my curiosity about how my choices are fitting into the pilot population at large, so I’ve decided to see what you all have to say about it! Since you can’t add a poll to an existing thread, I’ve set up a poll here: https://www.homebuiltairplanes.com/...54-design-study-candidate-configurations.html

    Let me know what you think!
     
    Last edited: Nov 28, 2015
    flyvulcan, billyvray, Himat and 2 others like this.
  13. Dec 6, 2015 #53

    Topaz

    Topaz

    Topaz

    Super Moderator Staff Member

    Joined:
    Jul 30, 2005
    Messages:
    13,727
    Likes Received:
    5,407
    Location:
    Orange County, California
    And the Winner is…

    DS54-C1-Side-Exterior.jpg
    DS54-C1-Plan-Exterior.jpg
    DS54-C1-Side-Interior.jpg
    DS54-C1-Plan-Interior.jpg


    Candidate 1

    Click on each image to enlarge

    Having dithered back and forth for a couple of weeks, I have finally made the decision to adopt Candidate 1 as the airplane I’ll continue to develop in this project. I know a lot of you really liked the other two, and obviously so do I, but “there can be only one”, as the movie line goes. Here’s how I arrived at my decision.

    Downselect 1
    First to go was Candidate 3. While this did the best in the polls, in the end I just couldn’t accept the idea of my fragile body being in a direct line between an impact point on the ground and 102 pounds of engine. I also didn’t like the fuel accommodation on this design. The fuel tank is directly under the engine and, even if protected by a horizontal firewall, anything that catches fire on the engine and drips down will be headed right for the fuel tank, on the outside of the airframe if not directly.

    As was noted in the poll thread once or twice, the aerodynamics in the wing-root/cowling area are somewhat risky. Many attempts at this configuration have had issues with flow separation in that area. While I think my solution of using the cowling as a bullet fairing was a good one, I don’t know for sure, and couldn’t really know until the airplane was finished and flown.

    All in all, Candidate 3 just had too many potential issues for my taste.

    Final Downselect
    This one was a very hard decision. I’ve been in love with my BV-141-based DS27 configuration for years, and someday I’m going to build that airplane. Somebody has to build that airplane.

    However, if it’s going to be designed and built, it ought to be done right. It really ought to have full retractable tricycle gear, a beautiful big custom canopy, and an uncompromising level of attention to aerodynamics and aesthetics. It would be best with some of the more-sophisticated composite structural methods we debated in the discussion thread. But the severe cost and simplicity restrictions I’ve placed on this project just won’t allow all that.

    In the end, the decision came down to one about “scope-creep”. It’s extremely easy to let a bunch of little changes, that make the airplane "better", completely blow a project out of the realm of reality. You start with a Piet-like concept and end up with something more like the Lancair IV. I'm sure I'm not the only one who has experienced this.

    So do I blow the budget and timeframe making the “awesome” airplane that might take another few years to develop, or do I stay on-target and build a “fun and good enough” airplane and have it a lot sooner and more affordably? Scope creep is a killer. I’ve battled it ever since I started drawing airplanes, back in my teens. This time, I’m sticking to plan.

    Another factor that played into my final decision was stated very well by DeepStall in the poll thread:

    That same sentiment has been nagging at the back of my head ever since I added Candidate 2 to the final list. Is it really wise to reach for the Moon the first time out? One of the goals I listed at the start of this project was to keep it simple, for this to be a good learning experience for me of running a design all the way through to a build, should I take it that far. Again, despite my strong desire to take a crack at Candidate 2, it's a very large bite to take at my first attempt.

    All of this together means Candidate 2 will have to wait for another day, when I can really do it justice.

    I should make clear that I'm not at all disappointed with Candidate 1. While the visibility isn't the best of the lot, obviously it's "good enough" in similar airplanes like the RF-4, and in nearly every other way, Candidate 1 is the simplest, easiest, most-manageable of the three designs. Performance should meet all my specifications, and this is going to be a heck of a lot of fun to fly! :grin:

    Next Post: I’ve spent a lot of time on this decision, and so progress on the project has stalled. Now that I know what it’s going to look like, overall, it’s time to get the numbers rolling again!
     
    billyvray, PTAirco, mcrae0104 and 6 others like this.
  14. Feb 3, 2016 #54

    Topaz

    Topaz

    Topaz

    Super Moderator Staff Member

    Joined:
    Jul 30, 2005
    Messages:
    13,727
    Likes Received:
    5,407
    Location:
    Orange County, California
    Interlude...

    I'm feeling very badly (for myself, and for anyone still following this thread) that I haven't been able to make any progress here since early December. This project is not done. I'll try to get another post up soon.
     
    RPM314 and bmcj like this.
  15. Feb 23, 2016 #55

    Topaz

    Topaz

    Topaz

    Super Moderator Staff Member

    Joined:
    Jul 30, 2005
    Messages:
    13,727
    Likes Received:
    5,407
    Location:
    Orange County, California
    Choosing an Airfoil
    For those of you wondering, “When the heck is this guy going to choose an airfoil for this thing?”, that time is now. When I started learning about airplane design, I did what everyone else does when they start that process: I wanted to pick the “right” airfoil first, and then I could figure out how the airplane would perform from there. Like so many things I’ve learned during this process, I’ve come to believe that this notion is backwards. You choose an airfoil to suit the airplane, not the other way ‘round. There isn’t any “right” airfoil for an arbitrary class of airplanes - sportplanes, gliders, etc. - that will be “best” for any airplane in that class, although airplanes within each class are often close enough in performance numbers that you see the same airfoils crop up within the class.

    I’ve finally learned enough about the DS54 that I can start fleshing out what I’m looking for in an airfoil. In general, I’m looking for a ‘foil that has quite low drag (meaning large extents of laminar flow - this is motorglider, after all), has been designed to accommodate plain flaps, and is about 12-15% thick - a good compromise between weight, high lift, and drag.

    Now let’s look at drag. Laminar airfoils tend to have a range of lift coefficients - and angles of attack, since the two are related - where the laminar flow is most extensive, and drag is at a relative minimum. This is usually called “the drag bucket” or “laminar bucket”. In the hypothetical example below, the Cl of an airfoil is plotted on the vertical axis against the drag coefficient on the horizontal axis as a blue line. (This graph is called an “airfoil polar”).

    Drag-bucket Graphic.jpg

    The idea is to choose an airfoil that meets all the other specifications above, and also provides a wide-enough drag bucket to keep the cruise Cl and the minimum-sink/maximum-rate-of-climb Cl inside that range of lift coefficients which has the lowest drag. If the wing were to be operating outside the drag bucket at cruise or minimum-sink/maximum-rate-of-climb airspeed, drag goes up and performance suffers. (For sportplanes, especially ones with larger engines, having the minimum-sink/maximum-rate-of-climb performance point inside the drag bucket is less important. Sportplanes usually just need a drag bucket large enough to accommodate a reasonable range of cruise speeds. Gliders and motorgliders operate engine-off over a very large range of lift coefficients, and airfoil selection is more difficult as a result.)

    Figuring out how wide a drag bucket I need means knowing the airfoil lift coefficient for this airplane for the two critical design points. Fortunately, I already know everything I need for these calculations.

    Next Post: Calculating the range of airfoil lift coefficients that define the "drag bucket" of the airfoil(s) I'll need, as well as the maximum lift coefficients flaps up and flaps down.
     
    RPM314 likes this.
  16. Feb 24, 2016 #56

    Topaz

    Topaz

    Topaz

    Super Moderator Staff Member

    Joined:
    Jul 30, 2005
    Messages:
    13,727
    Likes Received:
    5,407
    Location:
    Orange County, California
    Finding the Lift Coefficient Values I Need

    In order to choose an airfoil for my wing, I need lift coefficient values for four flight conditions:
    1. Stall speed, Flaps up
    2. Stall speed, Flaps down
    3. Minimum sink (maximum rate of climb)
    4. Cruise

    Further, I need these values to set up the “corners” of the flight envelope where I still want to be operating in the drag bucket. So I need these values at specific conditions:
    1. Stall speed, Flaps up - at maximum takeoff weight, ISO standard day at sea level
    2. Stall speed, Flaps down - at maximum takeoff weight, at the density altitude conditions in Table 1.
    3. Minimum sink - at the design weight of the aircraft after a climb to 3,000’ MSL, ISO standard day.
    4. Cruise - at the design weight of the aircraft at the end of the cruise portion of the design reference mission, 7,000’ MSL, ISO standard day.


    To make it easy, let’s bring up the design mission profile again, and run through the cases one-by-one:

    Design-Mission-Profile.jpg

    Stall Speed - Flaps Up (Vs[SUB]1[/SUB])
    If you’ll recall, I’m making an attempt to keep this airplane inside the Light Sport Airplane requirements, and that’s what set my flaps-up stall speed specification. It also sets the altitude conditions for meeting the specification: ISO sea-level, standard day.

    To figure out the lift coefficient of the wing under each condition, I’m going to use a variation of the basic lift formula:

    L = 0.5*ρ*V[SUP]2[/SUP]*S*C[SUB]L[/SUB]

    Where:

    L = Lift in pounds.

    ρ = Air density in slugs per cubic foot. (No, I’m not pulling your leg. The density of air in the SAE/Imperial system of units really is measured in “slugs per cubic foot”!)

    V = Airspeed in feet per second

    S = Wing area

    C[SUB]L[/SUB] = Wing lift coefficient

    To make the formula more useful for me here, I’m going to rearrange it so the answer comes out as the wing lift coefficient instead of the lift of the wing. I already know the lift of the wing, since in level flight it equals the weight of the airplane. So, a quick little bit of algebra later, the formula looks like this:

    C[SUB]L[/SUB] = (2*L)/(V[SUP]2[/SUP]*S*ρ)

    That done, let’s go back to the first case here - stall speed, flaps up.

    The LSA stall speed requirement is set in terms of flaps up, at 45 knots, which is 76 feet per second.

    What we really care about in terms of altitude is the air density, which is included in the formula. Looking up the ISO Standard Atmosphere chart in Appendix B in the back of Aircraft Design: A Conceptual Approach, I find that the air density at sea level on an ISO “standard day”, is ρ = 0.0023769 slugs per cubic foot.

    Wing area of my design is 98 square feet.

    The only remaining quantity is the lift of the wings, which equals the weight of the airplane. Note that I specified the weight as the maximum takeoff weight of the airplane, not the maximum gross weight. The airplane burns some fuel during start-up, warm-up, taxi, and a mag-check, and that weight is already gone by the time the airplane could ever be stalled. I’m kinda being pedantic about it in these first couple of cases, but the difference matters much more in the last two, and it’s a good habit to use the same methods throughout. Looking at the design mission diagram above, the airplane takes off at a weight of W[SUB]2[/SUB], which I know from my earlier sizing work is a ratio of 0.998*0.996 = 0.994 of W[SUB]0[/SUB]. So, at a maximum gross weight of 746 lbs., I get a takeoff weight at W[SUB]2[/SUB] of:

    Weight: W[SUB]2[/SUB]/W[SUB]0[/SUB] = W[SUB]0[/SUB]*0.998*0.996

    746*0.998*0.996 = 742 lbs.

    Like I said, I’m being very pedantic about process here - it’d be perfectly fine to just use 746 lbs. in this particular case.

    Now I’ll solve for the wing C[SUB]L[/SUB] at stall speed, flaps up.

    C[SUB]L[/SUB]= (2*L)/(V[SUP]2[/SUP]*S*ρ)

    = (2*742)/(76[SUP]2[/SUP]*98*0.0023769)

    = 1.10

    Okay, so that’s the lift coefficient of the entire wing at the flaps-up stall. What I want is the lift coefficient of the airfoil, so that I can look at airfoil polars and see if what I’m looking at meets my needs. Because this is a finite-span wing and not an infinite-span airfoil, the airfoil has to work a little harder to generate a given lift coefficient for the wing, making up for tip losses, etc. We could do all sorts of sophisticated analysis but, at this stage, Raymer uses a “close enough” fudge factor. Remember, my wing is going to change shape and size a little later on when I do the final optimization, so while I’ve accounted for the flight conditions, this really is “close enough” for now. Raymer says the lift coefficient of the wing is “close enough” to 0.9 of the airfoil lift coefficent, so I just need to divide my wing lift coefficent by 0.9 to get the number I want:

    C[SUB]l[/SUB]= 1.10/0.9

    = 1.22

    This means that I need to find an airfoil that, flaps up, develops a maximum lift coefficient of at least 1.22. Shouldn’t be too much of a problem.

    Stall Speed - Flaps Down (Vs[SUB]0[/SUB])
    Now let’s go through the same process, flaps down. When I decided that I needed flaps to satisfy my stall and take-off distance requirements (post #37), I recalculated my stall speed at the wing loading set at that time. Recall that this speed was for the much more difficult high-density-altitude conditions from Table 1 in my specifications sheet (Hemet-Ryan airport, on a 95°F day).

    The stall speed I calculated was 48 mph, which works out to 70 fps.

    The effective density altitude under these conditions is 4,089’ MSL. Appendix B in Raymer gives the air density in these conditions as ρ = 0.002111 slugs per cubic foot.

    The weight of the airplane is the same as before, at 742 lbs, and wing area is still 98 square feet (these are plain flaps, and don't increase wing area).

    Running the formula:

    C[SUB]L[/SUB]= (2*L)/(V[SUP]2[/SUP]*S*ρ)

    =(2*742)/(70[SUP]2[/SUP]*98*0.002111)

    = 1.46

    Taking that wing lift coefficient back to airfoil lift coefficient,

    C[SUB]l[/SUB]=1.46/0.9

    = 1.62

    So now I know that I need an airfoil that can generate at least a maximum lift coefficient of 1.22 flaps up, and at least a maximum lift coefficient of 1.62 using plain flaps. Progress!

    Next Post: Calculating the airfoil lift coefficient at minimum-sink and cruise.
     
    Last edited: Feb 24, 2016
  17. Feb 25, 2016 #57

    Topaz

    Topaz

    Topaz

    Super Moderator Staff Member

    Joined:
    Jul 30, 2005
    Messages:
    13,727
    Likes Received:
    5,407
    Location:
    Orange County, California
    Minimum Sink Airspeed
    Minimum Sink airspeed is that airspeed where the sink rate of the airplane, as a glider, is at the lowest value. Makes sense, right? Minimum sink is important to soaring for the case where thermals or other forms of “lift” are weak. If the airplane’s minimum sink rate is 100 fpm, it only takes a thermal going “up” at 100 fpm to stay aloft with no loss of altitude. By the same token, an airplane with a minimum sink rate of 150 fpm won’t be able to stay up in 100 fpm thermals - the best it can do is a 50 fpm altitude loss. No bueno. Conversely, if the thermal is roaring up at 800 fpm, the glider with 100 fpm minimum sink will be able to climb 50 fpm faster than the one that sinks at 150 fpm. Elevator … going up!

    Like the closely-related maximum climb rate, which is the same flight condition with an excess of power, minimum sink is most directly affected by drag. Every pound of drag is energy sucked out of the airframe, and gravity does the rest. So having the airplane operating in the airfoil’s drag bucket at minimum sink airspeed would be a plus. This defines the “high C[SUB]L[/SUB]” end of the drag bucket I need.

    The airspeed for minimum sink is usually almost exactly the same as that for maximum climb rate. The FAA has a handy generalization for this speed, as ~1.2*Vs. In my case, this would be the flaps-up stall speed, since minimum-sink flying will be done flaps-up.

    Vs[SUB]1[/SUB] = 76 fps

    V[SUB]V(min-sink)[/SUB] = 1.2*76 = 91 fps

    The altitude for all my soaring specifications is 3,000’ MSL, on an ISO Standard Day. Checking Appendix B in Aircraft Design: A Conceptual Approach shows that for these conditions, ρ = 0.002175.

    Weight for this case is a little more complicated. The flight condition presumes that I’ve started the engine, taxiied out, checked the mags, taken off, and then climbed up to 3,000’ MSL. At that point the engine is shut down and the prop feathered, and I start gliding. It’s fair to assume that the climb is goin to be done at full power. Let’s look at the design reference mission diagram and figure out the segments we need.

    Design-Mission-Profile.jpg

    I’ll be running the airplane from W[SUB]0[/SUB] through W[SUB]2[/SUB] (takeoff), and then about half-way up to W[SUB]3[/SUB]., which latter is at my design 7,000’ MSL cruising altitude. In fact, let’s just call it half-way and that’s close enough for today, and I’ll use the notation “W[SUB]X[/SUB]” for the 3,000’ MSL point half-way between W[SUB]2[/SUB] and W[SUB]3[/SUB].

    Pulling the weight fractions from posts #26-33, I have:

    W[SUB]1[/SUB]/W[SUB]0[/SUB] = 0.998
    W[SUB]2[/SUB]/W[SUB]1[/SUB] = 0.996
    W[SUB]3[/SUB]/W[SUB]2[/SUB] = 0.992

    Half the fuel fraction for W[SUB]3[/SUB]/W[SUB]2[/SUB] = 0.996. So, ...

    W[SUB]X[/SUB]/W[SUB]0[/SUB] = W[SUB]0[/SUB]*0.998*0.996*0.996

    = 746*0.998*0.996*0.996

    = 739 lbs.

    Wing area, as always, is 98 square feet.

    C[SUB]L[/SUB] =(2*L)/(V[SUP]2[/SUP]*S*ρ)

    =(2*739)/(91[SUP]2[/SUP]*98*0.002175)

    = 0.84

    Taking the airfoil lift-coefficient from this wing lift coefficient, I get:

    C[SUB]l[/SUB] = 0.84/0.9

    = 0.93

    What this means is that I know I need an airfoil whose drag bucket, at the high-C[SUB]l[/SUB] end, extends up to a lift coefficient of 0.93. That may or may not be a challenge. We’ll have to see what’s available.

    Cruise Airspeed
    The last lift coefficient I need is that of the cruise case. It should be pretty obvious that I’d like the airplane to cruise with as little drag as possible, so having the wing airfoil operating “in the bucket” in this condition is a given. Let’s figure this last one out. The cruise case is at a very high speed in the entire envelope, so this case defines the “low C[SUB]l[/SUB]” end of the drag bucket.

    The airspeed is my goal specification: 132 mph [115 knots] (194 fps). If it turns out the airplane isn’t this fast (it’s entirely possible), then that’s okay for our purposes here - if the airplane is flying more slowly, the lift coefficient for the same flying weight will be higher, and further into the drag bucket. I don’t expect the airplane will be faster than this, so my goal cruise speed is the right choice.

    The altitude for this case is also given by the specs: My design cruising altitude of 7,000’ MSL, ISO Standard Day. Referencing Appendix B again gives ρ = 0.001927.

    For weight, I want to keep the wing in the drag bucket throughout the design mission, so that means I’ll need the weight at the end of the cruise segment of the mission: W[SUB]4[/SUB] on the reference mission diagram above. Multiplying up all the mission weight segments as before:

    W[SUB]1[/SUB]/W[SUB]0[/SUB] = 0.998
    W[SUB]2[/SUB]/W[SUB]1[/SUB] = 0.996
    W[SUB]3[/SUB]/W[SUB]2[/SUB] = 0.992
    W[SUB]4[/SUB]/W[SUB]3[/SUB] = 0.955

    W[SUB]4[/SUB] = W[SUB]0[/SUB]*0.998*0.996*0.992*0.955

    = 746*0.998*0.996*0.992*0.955

    = 703 lbs.

    Plugging everything into the formula:

    C[SUB]L[/SUB] = (2*L)/(V[SUP]2[/SUP]*S*ρ)

    = (2*703)/(194[SUP]2[/SUP]*98*0.001927)

    = 0.20

    C[SUB]l[/SUB] = 0.2/0.9

    = 0.22

    So the low-C[SUB]l[/SUB] end of the drag bucket of my desired airfoil should be at 0.22 or less.

    Now I’ve pinned down quite a lot about this mystery airfoil for which I’m looking! For example, on a subjective level, I know I want it to be a “laminar” airfoil, with low drag for good soaring. I know that I need it to have a no-flaps C[SUB]lmax[/SUB] of at least 1.22, and a C[SUB]lmax[/SUB] with plain flaps of at least 1.63. I need it to have a drag bucket that ranges across lift coefficients from 0.22-0.93.

    Anything else before I go looking for candidates? That’s the subject of my next post.
     
    Last edited: Feb 25, 2016
    Himat, RPM314 and Pops like this.
  18. Feb 27, 2016 #58

    Topaz

    Topaz

    Topaz

    Super Moderator Staff Member

    Joined:
    Jul 30, 2005
    Messages:
    13,727
    Likes Received:
    5,407
    Location:
    Orange County, California
    Airfoil Selection (Continued)

    Three more discriminators for airfoil selection follow. I'm providing the background material here, and these will all fit into the matrix of airfoils and characteristics that I'll draw up for final selection. No math today.

    Pitching Moment Coefficient - Trim Drag
    One of the many forms of drag on a flying airplane is trim drag. Most airplanes balance out the forces of flight with a down-load on the horizontal tail. Canard airplanes, with the “tail” forward of the wing, do it with an up-load on the canard. There’s a price to pay for this tail lift, of course. For both conventional and canard aircraft, the lift produced by the tail creates induced drag - any surface creating lift makes induced drag in the process. A conventional airplane’s tail, “lifting” down, adds that force to that the wing has to carry, meaning it creates some additional induced drag as well as that produced by the tail. Having to carry the extra load, the main wing has to fly at a slightly higher angle of attack, which means a little more parasite drag as well.

    In a poorly-designed airplane, trim drag can add up to a noticeably high value, sapping performance out of the airplane throughout the flight envelope. For a sailplane, too much trim drag can be a disaster.

    A trim load on the tail is a consequence of positive static stability, so there’s only so much we can do to minimize the component that comes from balancing the CG ahead of the neutral point. However, another big source of trim drag is the pitching moment of the wing’s airfoil. Most airfoils produce a nose-down pitching moment when moving through the air, a result of the differing pressures across their surfaces. The closer to zero the pitching moment, the less load the tail has to carry in order to balance out that force, and less trim drag results. This is a good thing.

    Unfortunately, laminar airfoils very often have large pitching moments, on the order of -0.15 or even -0.20. A long tail boom can mitigate this - and you see that a lot on sailplanes. In my case, the length of my airplane’s tail is rather severely limited by the “standard” garage in which it will be built and stored. What that means is that a large factor in my final airfoil selection is going to be favoring the choice with the smallest pitching moment coefficient. I’d like to pick one that’s well under -0.15, if possible.

    Stall Characteristics
    I know some of you may be wondering why airfoil stall characteristics are so far down in this discussion. Stall characteristics tend to dominate most discussions of airfoil selection here at HBA.

    The fact of the matter is that airfoil stall characteristics play very little part in determining wing stall characteristics. The planform, twist, and spanwise airfoil selections of a real-world 3D wing can make a “vicious stall” airfoil into a pussycat of a wing, and vice versa. The balancing act is that the 3D wing adjustments to tame a sharp stall can sometimes reduce the efficiency of the wing, or limit the range of lift coefficients at which the wing can operate efficiently. Some of them can also improve the efficiency of the wing, too, so the designer chooses a strategy that provides maximum benefit at minimum penalty.

    Where this enters the picture in airfoil selection is in the gross overall stall characteristics of an airfoil. An airfoil showing a very sharp stall characteristic will probably need more “tweaking” in the wing to make sure there isn’t a wing-drop at the stall. I’d like to avoid that, so I’ll be looking for airfoils that have a soft, gentle stall to begin with. However, if I find an otherwise perfect airfoil, better than the other candidates in all ways except stall behavior, I’ll know that I can still choose the “sharp stall” candidate and tailor the wing geometry to give the entire wing a nice, soft, stall. Once I have a final "winner" in airfoil selection, I'll move on to tailoring the wing's stalling characteristics, so that I know they're safe.

    Cruise Airfoil L/D value
    Here’s another characteristic that tends to dominate discussions of airfoils, but that I’ll be using only as a “tie-breaker” in choosing between candidates.

    The L/D value of the airfoil at the cruise lift coefficient of the wing (and, additionally in my case, the maximum L/D of my airplane in soaring flight) will have an effect on the performance of the airplane in those conditions. All else being equal, I’d prefer an airfoil that has less drag for a given amount of lift at these two flight conditions, so if I have two remaining candidates which are otherwise much the same, the one with the better L/D value at these two flight conditions will be the winner.

    Next Post: Reynold's Number operating range.
     
    Last edited: Feb 27, 2016
  19. Mar 4, 2016 #59

    Topaz

    Topaz

    Topaz

    Super Moderator Staff Member

    Joined:
    Jul 30, 2005
    Messages:
    13,727
    Likes Received:
    5,407
    Location:
    Orange County, California
    Reynold’s Number Operating Range
    Air, like motor oil or maple syrup, is actually a little viscous - it “flows” and sticks to things as it flows over them. Obviously air is a lot less viscous than motor oil or syrup, but the effect is there nonetheless. The effect of this viscosity is dependent upon several factors including the air temperature, and inertial effects of the air itself such as the length of the surface over which the air is flowing, and these relationships between viscous and inertial forces are expressed as the Reynold’s Number. I’m not going to go into this in detail, so if you want to know more check it out with Google. But real-world effects of Reynold’s Number influence how much laminar flow an airfoil can sustain, how much lift it will produce, and can even change the stall characteristics of the airfoil. It’s important for me to know the Reynold’s Number under which my wing airfoils will be operating, at each flight condition of interest, so that I can use an appropriate polar for that airfoil under those conditions.

    Recall that, above, I mentioned that the length of the flow path changes the Reynold’s number. My wing is rather highly tapered, which means that the flow path at the root is much longer than that at the tip. For that reason, the tip will be operating at a lower Reynold’s Number than the root for any given flight condition. I’ll need to account for that.

    In the ‘Designing Your Homebuilt’ series of articles for Sport Aviation, John Roncz provides a handy spreadsheet for calculating Reynold’s Number. You can download a copy of that spreadsheet here: https://app.box.com/s/75rkq0tk4gne4wild1aim6upnd54v0fq

    What I’ve done is to set up a table of particular airspeeds and flight conditions that are of concern to me in choosing an airfoil, and listed them in columns. Then I measured the chord of my airplane’s wing at the root, at the tip, and at the taper break between them, and set those up in rows. Finally, I ran the numbers through the spreadsheet and wrote down the results for each combination.

    48 mph/4089'MSL@95°F 52 mph/Sea-Level ISO 61 mph @ 3500'MSL ISO 77 mph @ 3500'MSL ISO 132 mph @ 7000'MSL ISO
    BL Station* Chord Flaps Down Stall Flaps Up Stall Min. Sink Glide Best L/D Glide Cruise
    0" 35.9" 1,100,000 1,400,000 1,600,000 2,000,000 3,100,000
    151" 28.8" 900,000 1,200,000 1,300,000 1,600,000 2,500,000
    248" 15" 500,000 600,000 700,000 800,000 1,300,000
    As you can see, the values range from about 3,000,000 at the root at cruise, all the way down to about 500,000 at the tip when the airplane is stalling, flaps down. That’s a very large range, but fortunately I don’t need full airfoil performance across all of it. In fact, while I need the Cl[SUB]max[/SUB] performance to extend all the way down to a Rn of 500,000, I only need drag performance (width of the drag bucket) down to the minimum-sink airspeed, or about 700,000.

    Section Drag
    Earlier on, I said that I wanted a “laminar” section for this airplane. That’s because this is a motorglider, and I need the low drag for soaring. But how low drag? It’s actually possible to sit down and calculate how low the airfoil’s C[SUB]d[/SUB] needs to be in order to meet all the performance requirements, but the error bars on those calculations are really far apart with the tools I have at my disposal, so I’m not going to go there. Instead, I looked at a bunch of airfoil polars already in-use on production sailplanes, and arrived at a desired “ceiling” for the airfoil C[SUB]d[/SUB] inside the drag bucket. To give myself some wiggle room and acknowledging that this isn’t a competition racing sailplane, I’m going to say that I want an airfoil that has a drag-bucket C[SUB]d[/SUB] of 0.008 or less. Much less would be even more desirable.

    Putting It All Together
    Looking at all these requirements I’ve developed, I now have enough information to describe the airfoil for which I’m looking:
    • 12-15% thick
    • Suitable for use with plain flaps
    • A negative Cm[SUB]zero[/SUB] >= -0.1 (no flaps)
    • Cl[SUB]max[/SUB] (no flaps) of at least 1.22, at a Rn 600,000-1,500,000
    • Cl[SUB]max[/SUB] (plain flaps) of at least 1.63, at a Rn of 500,000-1,500,000
    • Drag bucket extending from C[SUB]l[/SUB] = 0.22-0.93, at a Rn operating range of 700,000-3,000,000
    • C[SUB]d[/SUB] < 0.008 throughout the drag bucket at a Rn operating range of 700,000-3,000,000

    The Hunt Begins!
    Now it’s time to go hunting for this airfoil. Does it exist, or will I have to resort to other aerodynamic tricks to make it all work? I have my books, I have Airfoiltools.com, and I have … you! Do you know an airfoil that meets these specifications? Really meets these specifications? Tell me about it in the discussion section. All I ask is the following:
    .
    1. Your selection really meets the specifications I've listed.
    2. You have (or can point me at) actual coordinates for the section.
    3. You have a polar that shows that it at least nominally meets the specs at some given Reynold’s Number.
    4. It’s a section I can actually use: It’s not some proprietary section for which I’d need a license.

    * "BL Station" means "Buttline Station", a reference to the standard coordinate system for describing airplanes (and boats, from which we stole the concept shamelessly). From a datum point, any point X,Y,Z on the airplane can be measured as "Station X", "Buttline Y", and "Waterline Z". "Station" is front to back, "Buttline" is out along the wing from the centerline, and "Waterline" is up from the bottom (makes more sense with boats). So "BL Station 248" on my airplane is 248 inches from the centerline of the aircraft - which corresponds roughly to the wing tip. BL0 is the wing root at the centerline of the airplane.
     
    Last edited: Mar 22, 2016
  20. Mar 16, 2016 #60

    Topaz

    Topaz

    Topaz

    Super Moderator Staff Member

    Joined:
    Jul 30, 2005
    Messages:
    13,727
    Likes Received:
    5,407
    Location:
    Orange County, California
    Candidates
    Looking at the requirements list for my airfoil, I see that I’m looking for a highly-laminar airfoil with moderate camber, with that camber somewhat forward rather than way back towards the aft half of the airfoil. Older laminar sections tended to have aft camber because it helps promote long laminar flow runs, but the price to pay was a high to very high pitching moment, since it means the airfoil develops most of its lift well-aft on the chord. Much of the modern work on airfoils has been to “fix” this. My airplane has a rather short tail (by sailplane/motorglider standards), so I have to be very careful about the pitching moment of the airfoil I choose. A high-moment airfoil would lead to a lot of trim drag.

    With this knowledge in hand, I looked for candidate airfoils in a number of ways: suggestions from HBA members, recommendations from books, and by comparing similar airfoils in Airfoiltools.com. In no particular order, here are the ones at which I’ve decided to look further:

    fx79k144-il_l.jpg
    Wortmann FX 79-K-144/17 - This is a sailplane airfoil from Dr. Wortmann, who is famous for his many airfoils in that field. This one was suggested by Autoreply here on HBA, as a more-modern alternative to the older Wortmann FX 67-K-150/17 I was considering earlier. The aft camber on this one is really odd, as if it has a permanently-reflexed flap. That's an obvious attempt to keep the pitching moment down. (The camber line of each airfoil is the green line running lengthwise half-way between the upper and lower surfaces.)

    r1082t-il_l.jpg
    Roncz 1082T - This airfoil was used on the outer main wing panels on the world-circling Voyager aircraft. It was designed by John Roncz, who did the airfoils for many of Burt Rutan’s designs. I stumbled on this one in a search of similar airfoils on Airfoiltools.com. While it’s not explicitly a sailplane airfoil, it shares a lot of characteristics with them, and I know it was engineered to preserve much of its performance even when contaminated with bugs, rain, or dirt. At 16%, it's a bit thicker than the other candidates, which might be a concern for me with my full-depth foam wing.

    e662-il_l.jpg
    Eppler 662 - Dr. Eppler’s 662 ‘foil is another sailplane airfoil, which was highly recommended in Alex Strojnik’s book, Laminar Aircraft Design. However, it’s pretty old, so I’m curious how it will hold up against more-modern designs. Notice the heavy aft camber.

    e642-il_l.jpg
    Eppler 642 - Another sailplane airfoil by Dr. Eppler. I found this one when I was looking for the Eppler 662 above. I know nothing else about the history of this selection, but it’s performing well in XFoil.

    GA40415.jpg
    Riblett GA40415* - Harry Riblett’s book of airfoils is one of the better gifts we’ve gotten in the sport aviation community. They “fix” some issues with the NACA 6-series airfoils, keeping the good characteristics and improving stall behavior, among other things. The Riblett 40-series airfoils are the most-laminar, lowest-drag of the entire family, and I’ve chosen a middle-range camber example, with about the same camber as my other candidates.

    That’s my list. I’m running these through XFoil, developing a table of numbers that will let me look at them together and make a final selection. I’ll talk more about the process next time.

    ----------
    * You'll note that the airfoil in the image is listed as "GA40A415". Riblett's laminar airfoils come in two variants, one with a slight upper- and lower-surface "cusp", or narrowing of the afterbody (which you can just see with a dashed line in the drawing), and the an "A" variant without the cusp. The cusp reduces drag a little further, but can make for heavier aileron forces because of the narrow angle between the upper and lower surfaces near the trailing edge. In my case, I want the drag reduction, and my narrow-chord ailerons aren't likely to produce very high forces, even with the cusp.
     
    Last edited: Mar 16, 2016
    bmcj, fly2kads and steveair2 like this.

Share This Page

arrow_white