Question(s) about Steel Tube Fuselage Design

Discussion in 'Aircraft Design / Aerodynamics / New Technology' started by Tom Kay, Jan 21, 2010.

Help Support HomeBuiltAirplanes Forum by donating:

  1. Jan 21, 2010 #1

    Tom Kay

    Tom Kay

    Tom Kay

    Well-Known Member

    Joined:
    Aug 10, 2007
    Messages:
    398
    Likes Received:
    15
    Location:
    Ottawa Canada
    Hi Again;

    I could have put this in the tube and rag section, but it is about design principles, so I chose the Design section. Forum chiefs, move it if you prefer.

    So, I would like to know if it's acceptable practice to design a welded 4130 steel tube fuselage with different sized "cells" (as viewed from the side). What I mean by "cells" is the open areas between the vertical side tubes. Of course, what I'm really talking about is spacing the verticals differently at different locations from nose to tail.

    Around the wing area, where it attaches to the fuse, you've got a lot going on. There might be 2 or 3 attach points (indicated by arrows), and I would never have them bolt onto the side of a tube wthout having a vertical or diagonal meet at that location. This is why we might need more vertical tubes added into certain areas of the truss structure.

    The best way I can explain what I mean is to show a simple picture. It's a side view of a 3/4 scale Mustang replica, but it could be any other similar aircraft. Notice that there are simply more tubes in the area that's circled.

    So, is this acceptable practice, or do the stresses and forces not "flow" properly with differently spaced tubes like you see in the picture?

    Thanks, Tom.
     

    Attached Files:

  2. Jan 21, 2010 #2

    Tom Kay

    Tom Kay

    Tom Kay

    Well-Known Member

    Joined:
    Aug 10, 2007
    Messages:
    398
    Likes Received:
    15
    Location:
    Ottawa Canada
    And before you get a chance to catch your breathe, here's a follow-up question or two. I wanted to separate these, as they deal with different aspects of tube fuse design.

    I know that if I add a ton of additional vertical tubes, it will add weight. That's bad. But is there anything fundamentally wrong with more and smaller "cells" in the tube structure? Does this in any way weaken the whole truss assembly? Does it improve strength overall?

    Notice too, that I have vertically flipped the diagonal members of the bottom picture, compared to the top picture. You might not see that if you just glance quickly at the images, but look at the arrow that shows where the front spar attaches. I had a reason for doing that.

    Let's say my pilot skills are a bit rusty and I have a hard landing. The biggest arrow represents a hard landing. I was thinking that with the diagonals arranged as they are in the bottom picture, the hefty forces that are felt in the landing gear, then the spar, then the fuselage, would be withstood better if there were 3 structural members "sharing" the load, as indicated by the 3 arrows I've drawn. Simply asked, is this a better design than the top picture, based on heavy landings?

    Thanks again, Tom.
     

    Attached Files:

  3. Jan 21, 2010 #3

    DaveK

    DaveK

    DaveK

    Well-Known Member

    Joined:
    Apr 22, 2007
    Messages:
    343
    Likes Received:
    17
    Location:
    Northern California
    More experienced types on here will undoubtedly chip in but to answer your first question there really isn't any set "rule" about what you call cell spacing. The truss members are all sized individually to handle the calculated loads which are more critical in compression due to buckling. In general you have a trade off of increasing the number of truss members which shortens them and makes them more buckle resistant against more construction hours needed to fit and weld them together. Also if a tube gets thin enough it gets very hard to weld properly. So as always its a trade off during design.
    So you have to weigh that trade during the design of the truss and you can come up with several different trusses that can handle the loads, but they all won't be the same weight. I wouldn't jump to the conclusion that adding additional vertical members will increase the overall weight of the truss. As I said the increased number of truss members tends to increase the buckling resistance of the truss members due to them being shorter and the individual truss members end up thinner then in the first truss. You will have to calculate the weight for each completed truss to know, your eyeball won't help here.
     
  4. Jan 22, 2010 #4

    Tom Kay

    Tom Kay

    Tom Kay

    Well-Known Member

    Joined:
    Aug 10, 2007
    Messages:
    398
    Likes Received:
    15
    Location:
    Ottawa Canada
    Hi Dave;

    Some good points, thanks, especially about not assuming that more vert members will automatically increase overall weight.

    While I'm on this topic, I wonder, does anyone know of a book that deals largely with steel tube fuselage design? I just bought on on sheet metal working, at the suggestion of Mr. AVI, but having a similar book on truss structures for aircraft would be good.

    I just had a quick look on Google for such books, and there are some on technique, but I'd be more into learning about the design principles.

    Thanks, Tom.
     
  5. Jan 22, 2010 #5

    djschwartz

    djschwartz

    djschwartz

    Well-Known Member

    Joined:
    Jun 22, 2008
    Messages:
    982
    Likes Received:
    94
    Location:
    Portland, Oregon
    A couple of quick comments.

    Take a look at the fuselage trusses of a Bucker Jungmeister and a Pitts Special. Bucker used a large number of small tubes, Pitts used a smaller number of larger tubes. Both resulted in excellent aerobatic biplanes. Biggest difference, the Pitts is easier to build.

    Unless you do something really silly the truss will be a very small fraction of the weight of the aircraft. The truss for the Pitts S1 weights about 30 lbs. The truss for my Stephens Akro, about 70 lbs. I've had the airplane disassembled down the the bare truss and I can easily lift it myself.

    Truss design is somewhat iterative, as is most aircraft design. Most engineering of all disciplines for that matter. Start by looking at the truss structure of some aircraft similar in concept to the one you propose. Lay out a truss and add your loads. Look at the resulting stress in each member and see that that would mean for the sizing of that member. A lot of the details will be determined by where you want to place the point loads, spars, pilots, fuel, etc. You don't want any of these loads transferred to the middle of an element, they need to be carried to a junction.

    Most basic physics or mechanical engineering textbooks contain a section of truss stress calculation methods. it's pretty basic and can even be automated with a spreadsheet. Understanding what the loads are can be a bigger challenge than calculating the stress in the truss.

    Dave
     
    Last edited: Jan 22, 2010
  6. Jan 22, 2010 #6

    Mac790

    Mac790

    Mac790

    Well-Known Member

    Joined:
    Jun 22, 2008
    Messages:
    1,529
    Likes Received:
    19
    Location:
    Poznan, Poland
    Tom,

    I also think, that it might be a good idea to invest some $ in some plans. For example plans for aerobatic monoplanes like One Design 107, Laser Z200/Z2300 or maybe even for Sonerai might be very useful. Those plans aren't expensive (compared to real aircraft costs).
    Second, I agree with Dave, it could be a good idea to get some book for stress analysis, special text book for undergraduate students, those books are rather simple to understand, and usually besides theory you can find some examples of calculations there.
    Third, next try to get Femap demo, PTAirco uses it in his designs with success.

    Check out also those pictures for a spar attachment ideas.

    Seb
     

    Attached Files:

  7. Jan 22, 2010 #7

    Tom Kay

    Tom Kay

    Tom Kay

    Well-Known Member

    Joined:
    Aug 10, 2007
    Messages:
    398
    Likes Received:
    15
    Location:
    Ottawa Canada
    Dave;

    Thanks for the reply. I keep convincing myself that I'm not up to stress analysis or load calcs, but maybe I shouldn't be so defeatist. I suppose I could start with a few simple examples, and see how it all works. I do, however, recall that Stress Analysis was my worst subject in Aerospace classes, 30 years ago. Ugh. But I'll also hunt down views of the Jungmeister and Pitts S1.

    Seb, I recall you showing me the Laser a while ago, and by coincidence, I just found it today again. Great examples, but I have decided that I will have my fuselage frame completely mounted above the wing, not bolted through the spar like the Laser. However, I will closely look at it again just to get ideas from the tube layout.

    Appreciated as always. Tom.
     
  8. Jan 22, 2010 #8

    Mac790

    Mac790

    Mac790

    Well-Known Member

    Joined:
    Jun 22, 2008
    Messages:
    1,529
    Likes Received:
    19
    Location:
    Poznan, Poland
    Tom,
    I remember that, but it's One Design 107 this time, it's low wing monoplane.
    [video=youtube;VF6IkFTwCmA]http://www.youtube.com/watch?v=VF6IkFTwCmA&feature=related[/video]
    I don't know why are you going to make it this way, almost everybody else (Extra, Zivko, Laser, even Russians check out attachment, etc), bolt it through the spar.

    Seb
     

    Attached Files:

    Last edited by a moderator: Mar 20, 2019
  9. Jan 22, 2010 #9

    Tom Kay

    Tom Kay

    Tom Kay

    Well-Known Member

    Joined:
    Aug 10, 2007
    Messages:
    398
    Likes Received:
    15
    Location:
    Ottawa Canada
    Hi Seb;

    Thanks again for the reply. I was considering the other day why I couldn't use the bolt-through approach for attaching the main spar for a 3/4 scale Mustang. Something made me think that this would not work on the P-51, specifically. And it finally occurred to me, although it might be immediately obvious to others.

    Take a look at how the landing gear folds on a Mustang, and see where the wheels end up when retracted. Notice that they are actually well inboard of the fuselage frame sides. If we had the lower longeron of the frame near the bottom of the fuselage, in the wing attach area, then the main gear would smack into the lower longerons each time they attempted to retract. See what I mean?

    I was also curious why Titan had made their fuselage this shape but then it made itself apparent when I looked at the gear.

    Any thoughts around this problem? I can't think of any.

    Cheers Seb, Tom.
     

    Attached Files:

  10. Jan 22, 2010 #10

    djschwartz

    djschwartz

    djschwartz

    Well-Known Member

    Joined:
    Jun 22, 2008
    Messages:
    982
    Likes Received:
    94
    Location:
    Portland, Oregon
    No offense intended; but, if you don't feel you're up to load and stress calculations I wouldn't recommend designing your own fuselage from scratch. Stick to minor modifications of something proven. Otherwise you're likely to end up with something either excessively heavy or unsafe. Buy a set of plans for something similar to what you want to end up with as was suggested and work from there.

    Dave
     
  11. Jan 22, 2010 #11

    Tom Kay

    Tom Kay

    Tom Kay

    Well-Known Member

    Joined:
    Aug 10, 2007
    Messages:
    398
    Likes Received:
    15
    Location:
    Ottawa Canada
    Hi Dave;

    No offense taken, but I likely won't do it as you suggested. There is nothing really available that fits the bill, and Titan, for example, doesn't sell plans. I also don't like how their fuselage is constructed, at least not entirely.

    Also, I tend to push myself in certain areas. I can learn, and also benefit from colleagues at work, since we have so many stress and finite element guys and other aircraft designers, and then there are also online truss calculators. None of this is necessarily elegant or easy, but that's OK.

    Before I start welding, I would make sure, using all the resources I have (within reason) that I am building an acceptable design. Your point is well taken, but outside of what I'm after.

    Cheers, Tom.
     
  12. Jan 23, 2010 #12

    Autodidact

    Autodidact

    Autodidact

    Well-Known Member

    Joined:
    Oct 21, 2009
    Messages:
    4,513
    Likes Received:
    799
    Location:
    Oklahoma
    Schaum's outline of Strength of Materials has sections on truss design as well as buckling of columns (tube under compression), it is fairly easy to understand (or I wouldn't have been able to understand it!). Even if you don't know the calculus used in the derivations of the formulas (mainly the buckling problems), you can still use the basic formulas, I think, so long as you can do a bit of trigonometry.
    Aside from buckling, tensile/compressive stress is pretty simple, it's figuring out what those stresses are (force and direction) that's kind of tricky and thats where the trig comes in.
    Raymers explanation of motor mount design (Simplified...) is a good introduction to trusses.
     
    Last edited: Jan 23, 2010
  13. Jan 23, 2010 #13

    BoeveP51

    BoeveP51

    BoeveP51

    Well-Known Member Lifetime Supporter

    Joined:
    Jan 25, 2008
    Messages:
    97
    Likes Received:
    10
    Location:
    Spring Hill, FL - formerly W. Ossipee, NH
    Tom, you should take a look at the steel fuse plans from Jurca and Boeve. This would give you some insight. These have both been around for quite some time with Jurca having had a number of planes flown with the steel fuse design.
    These designs are way overkill in my estimation but then I am not an engineer.
    After looking at the Titan when it was first introduced I thought it was a toy compared to the Jurca and would break apart on the first flight. Guess my engineering eye was wrong on that account.

    Merle
     
  14. Jan 23, 2010 #14

    djschwartz

    djschwartz

    djschwartz

    Well-Known Member

    Joined:
    Jun 22, 2008
    Messages:
    982
    Likes Received:
    94
    Location:
    Portland, Oregon
    Tom, If you have resources like that available you should be able to come up with a good design. One of the advantages of Chrome-moly trusses is that you don't have to engineer to the n'th degree just to have a good result as long as you're not trying to do anything really strange. Good luck and have fun!

    Dave
     
  15. Jan 25, 2010 #15

    Tom Kay

    Tom Kay

    Tom Kay

    Well-Known Member

    Joined:
    Aug 10, 2007
    Messages:
    398
    Likes Received:
    15
    Location:
    Ottawa Canada
    Hi Guys;

    Autodidact; I dug out my dusty copy of Schaume's Materials and am starting to go through it. I couldn't find a huge amount on trusses just leafing through it, but I'll go through it more thoroughly to see. I also recall reading Dan Raymer's engine mount section, and it was pretty informative. One challenge is to get a guy like me, who's almost 30 years out of Aerospace school to recall any math, or to explain it in a way that a deep sea sponge would get. That's about what I need at this point. One slow step at a time.

    Boeve P51, good idea on the checking of other plans. I have done this with a few aircraft so far, although none of the replica fighters. Just a couple of biplanes. The Titan does look somewhat flimsy with its 1/2" square tubing, but the thing I didn't know at first, was that the skin does take some load, especially shear loads to resist twisting, and probably bending as well.

    I think I'd like to make my tube frame take almost all of the loads, with the skin just being cosmetic, and I don't feel that this would be too hard to do. Adding diagonal members will cut down a lot on the torsional flexing of the frame. Now I just have to figure out what kind of loads I'd be asking the frame to withstand, and where the loads would be applied.

    I did make a full sized frame mockup out of 3/4" square pine for a Murphy Renegade biplane. This is one I am still considering, and a realistic project for any beginner. The frame size is fairly close to the 3/4 scale P-51, and I was pretty surprised to see how much the frame stiffened up when I started adding diagonals. I haven't completely finished it, or tested it in any way, but I have learned a few useful things. However, at the end of the day, the lure of a P-51 replica is hard to ignore !

    Dave;

    I shouldn't "brag" that I have tons of resources at my disposal, because it's not like I snap my little fingers and FEA guys jump to it. But I do work in an aerospace research facility, and we do have some guys who like this sort of thing. That will help. And the software that I talked about might be of some help. I started to play with both CADRE and a freeware called 2D Truss Analysis. I'll see how that goes. If I can figure out the software, and if it really analyses loads of each member in a truss, can you imagine how useful that would be?

    Lastly, I suspect that you are quite right about not "gonculating" everything to the Nth degree with steel tube trusses. For one thing, you can't get wall thicknesses in increments of one thou, so you have to work with the next nominal size up, or just stock a few common ones for any given project.

    Thanks all, Tom.
     
    Last edited: Jan 25, 2010
  16. Jan 25, 2010 #16

    Autodidact

    Autodidact

    Autodidact

    Well-Known Member

    Joined:
    Oct 21, 2009
    Messages:
    4,513
    Likes Received:
    799
    Location:
    Oklahoma
    Tom, between your workmates and this forum you should be all right on the math; I know a little and there are others who are quite proficient. For me, it was understanding the concept of taking moments about a point (the trigonometry is relatively simple, but crucial). Don't be afraid to post up a reasonably neat sketch when asking a truss related question (pic worth a 1,000 words).
    Also, if you can use the software that figures the loads in trusses as well as software that figures the aerodynamic loads, you'll have the busy work knocked out :grin:.
    I just happen to be one of those deep sea sponges, myself!
     
  17. Jan 25, 2010 #17

    BoeveP51

    BoeveP51

    BoeveP51

    Well-Known Member Lifetime Supporter

    Joined:
    Jan 25, 2008
    Messages:
    97
    Likes Received:
    10
    Location:
    Spring Hill, FL - formerly W. Ossipee, NH
    Tom,
    I am attaching a JPG of my fuse. This is for a 2/3 scale P51. This is from the Jurca plans.

    Merle
     

    Attached Files:

  18. Jan 26, 2010 #18

    Tom Kay

    Tom Kay

    Tom Kay

    Well-Known Member

    Joined:
    Aug 10, 2007
    Messages:
    398
    Likes Received:
    15
    Location:
    Ottawa Canada
    Boeve P51

    Thanks for the pic of Jurca's P51 frame. The one thing I don't understand is how can he have a bolt-through spar (I think I see mounting plates for the spar) and tuck his wheels up? I must be missing some detail.

    Tom.
     
  19. Jan 26, 2010 #19

    BoeveP51

    BoeveP51

    BoeveP51

    Well-Known Member Lifetime Supporter

    Joined:
    Jan 25, 2008
    Messages:
    97
    Likes Received:
    10
    Location:
    Spring Hill, FL - formerly W. Ossipee, NH
    The wheel wells are outboard of the fuse. My Boeve wing and Thunder Mustang wing are both this way. Can send a pic if you like
    Merle
     
  20. Jan 26, 2010 #20

    Tom Kay

    Tom Kay

    Tom Kay

    Well-Known Member

    Joined:
    Aug 10, 2007
    Messages:
    398
    Likes Received:
    15
    Location:
    Ottawa Canada
    Hi Merle;

    Yes, any pics would be helpful. I'd like to have a look at the layout and positioning of the gear on these Mustangs. As I think I mentioned, designing a scale replica seems to be an ongoing series of compromises.

    I'm not sure whether you should just post them in this thread, or send them to my Private Message board. I seem to be completely forgetful to check my PM's every so often. However, I'll try to watch for any incoming messages.

    Thanks again, and within reason, feel free to send all the pics you want.

    Tom.
     

Share This Page



arrow_white