Reverse Engineering an Existing Commercial Design

[Nims11]

There are several great composites airplanes out there such as the Cirrus, Diamond, and Flight Design. All these planes new are well over $100k, I believe. I was wondering how difficult it would be to copy the design for the purpose of building ones own airplane for personal use. The raw materials for the airframe should only be a fraction of the total sticker price and then you could add the engine and instruments to meet your budget, I know some of these planes have very expensive instruments.

Would it be difficult to copy the design from existing data? Would you need the actual plane to somehow make a mold from? Is this a doable exercise for someone with skill in composite building?

Regarding the legality, yes I know it is not legal. But if I was to build this for myself, and not for sale, and since I could never afford to buy one otherwise, I don't have any moral reservation about doing this.

 

[BoKu]

There are several great composites airplanes out there such as the Cirrus, Diamond, and Flight Design. All these planes new are well over $100k, I believe. I was wondering how difficult it would be to copy the design for the purpose of building ones own airplane for personal use.

Consider that most RV-10 builds come in north of $100k, I seriously doubt you can make a nice Cirrus equivalent for less than that. To do it halfway decently, you'd need to make female molds for most of the large parts, and a lot of other parts as well. That's pricey, and you'd have to commit to making a bunch of units to pull the price down to what Cirrus charges.

And, yeah, you could do a one-off Cirrus-ish airplane using low-volume soft-tooled approaches like on the Rutan airplanes. But when you look at the time you've spent, you might have well have just bout a Cirrus.

...Regarding the legality, yes I know it is not legal...

Actually, it's not all that illegal, just kind of pointless.

Thanks, Bob K.

 

[Topaz]

...Would it be difficult to copy the design from existing data? Would you need the actual plane to somehow make a mold from? Is this a doable exercise for someone with skill in composite building?...

You'd basically have to obtain one that you can completely tear apart. These are composite airframes, and some of the structure is completely closed out to "normal" access. To get measurements, you'd have to cut into those closed-out areas. Plus you'd need to strip down all the composite layups layer-by-layer so that you can determine the layup schedule in each and every one. So you'd need to buy, and then destroy, a new (or used) $100k+ airplane in order to get all the structural dimensions for your "less-expensive" version.

And then comes the bigger problem: Production aircraft use production tooling, and production methods. Even if you were to duplicate the shapes and dimensions, you have to duplicate the structural qualities as well. To have the same structural qualities in the same dimensioned parts, you're going to need to use the exact same materials and methods as the production model - duplicating all their tooling and processes. And how do you find out exactly what resins and reinforcement they're using, and what production cycles they use in laying up their composites?

Barring that (and why would you do it?), you're going to need to redesign the structure from scratch for the materials and methods that you'll use yourself. And if you can do that, you can design exactly the airplane you really want, anyway, without having to copy someone else's design.

There really aren't any short-cuts. Buy a Cirrus/Diamond/Flight Design, buy an existing kit/plans, or design your own. Those are your realistic options. Copying airplanes like this really isn't among them.

 

[Dana]

I'm not familiar with its details, but like any modern aircraft, the Cirrus' design is tied up with its production techniques and equipment. For an amateur to duplicate it, he'd have to invest in equipment and tooling costing far more than a new factory aircraft. Alter the construction to suit amateur one-off techniques, you have a homebuilt which looks like a Cirrus and may have similar performance... why bother? It has the compromises inherent in any factory built and publicly marketed aircraft. Better to build a homebuilt design with better performance.

Dana

 

[Doug2233]

In my opinion: very unrealistic. See Topaz response...

How do you suppose you would identify and replicate the layup, foam type and thickness, cloth types, direction and matrix without having access to the design documentation? It might look simple enough and appear to be easily reproducible but even with specialist engineering experience it would be a very difficult and time consuming exercise. Never mind the tooling and set up costs.

Talk with a professional organisation who undertakes composite structural repairs to aircraft (gliders for example) and see the extent that they go to to replicate the original structure for even the smallest structural repair.

 

[Wingcrafters]

I would point out that while reverse engineering the structure probably isn't realistic, you could use the basis of the design for an aerodynamic head start. Copying the outside of the airplane from existing drawings would give you an airplane with known flight characteristics at least. You'd still be faced with deigning the structure, but you'd be one step ahead of a completely original design. Not sure if can even obtain accurate drawings for that though. You'd need to know the exact airfoil, incidences washout etc. Not something Cirrus is likely to sell you.

 

[TFF]

Cirrus stared out making homebuilts. Unless you have one to cut up and copy, you will have to design everything anyway. Lancairs, Glasairs, Wheeler Express, SX300s, VK-30s, White Lighting Take a lot of work with just a kit. It will not be cheap if you look at the whole effort.

 

[cheapracer]

Would it be difficult to copy the design from existing data? Would you need the actual plane to somehow make a mold from? Is this a doable exercise for someone with skill in composite building?

It's all quite easy really but of course it comes down to your personal skillz and those of people who can do some figuring for you either by computational design work or just solid experience.

There are some pretty good details of many designs out there to make up a basic 3D then slice it up to enable you to make a full sized model from foam sections to make a plug. 3D scanning is getting cheaper all the time also and of course that would get all your wing foil, AoA, etc, details in one hit. Get to a hangar and speak to some owners and tell them what you are trying to do and could you have a good look and get some measurements from their crafts. Also getting a wrecked fuse would be a big advantage to figure out the structural side of things and composite matrix layups used. Not like there isn't a Cirrus fuse or 2 laying around somewhere. Might get enough info/templates from the wreck to make a plug.

One advantage for you is that for some reason most aircraft manufacturers don't like to use mixed materials, eg; a fiberglass plane must be all fiberglass, aluminium must be all aluminium, etc, but you can use a mixture to suit the purpose to make your life easier such as a fiberglass fuse with aluminium wings and steel support structures.

Actually a (production) plane I bought recently has a fiberglass fuse, aluminium wings with steel tube tie ins, seems a logical layout to me especially for a one off homebuilt. Unless you seriously know what you are doing, better do the wings in aluminium because of the wealth of knowledge and designs to draw from out there. Easier to repair and maintain as well.

Also there's some very accurate 3D models out there already that you can start from such as here ..

Cirrus SR22 Airplane 3D Model .3ds .lwo .lw .lws - CGTrader.com



and they are easy to get sliced up to cut your foam slices to.

Here's a sample I just grabbed off the net and sliced up of a low poly Diamond to give you the idea ..







and an example of what you are trying to achieve ..








Similar methods are old news to boat builders, read the 7 pages here for example ..

http://www.bateau2.com/howto/foam1.php

 

[Nims11]

These replies are much appreciated, it gives me a better perspective on what all goes into building an airplane. And I love that this site has the diversity of knowledge where you can get one reply saying you basically have to take a Cirrus and tear it up, to another saying that the build is "quite easy".

Judging by some of the replies, I think my question must have been worded too narrowly. There are two aspects to replicating an airplane like the Cirrus that are being discussed in this thread. One aspect is from a composite building perspective; determining how to create the plane from composite techniques, figuring out how to build the molds and what techniques to use in laying up the fiber etc. The other aspect is from an engineering perspective; taking all the data known about the plane, dimensions, wing area, weight and CG, stall speed, etc. And using that data along with engineering tools to determine what structural requirements are needed and so on. Obviously both of these aspects are necessary to actual build the plane. So instead of taking a Cirrus and tearing it apart, which is a bit melodramatic, could you not use engineering tools coupled with all the known data about the plane to determine what kinds of strengths are needed and where, and therefore what type and amount and technique of fiber glass is needed, etc?

Engineers who have designed planes like the KR2, or the Long EZ, or whatever, basically start from a clean slate and have to design everything from the bottom up using all their engineering knowledge and tools. And yet they are never really going to know what that plane is going to do until they have actually built it and flown it. It seems like it would be a lot easier to take an existing design, given everything known about the plane, including its flight performance and characteristics, and using engineering tools and composite design techniques to create plans that would yield a homebuilt airplane that was pretty much identical to a Cirrus. I'm not suggesting that a designer should do this in lieu of creating an original design, if they are passionate about designing airplanes then they would obviously want to create there own original airplane. I was just wondering if someone with engineering and composite skills could fairly straightforwardly create plans for a replica of a quality airplane such as a Cirrus or Diamond, because I think from a homebuilders standpoint this would be useful.

Why would it be useful? Let's switch our example plane to the Diamond DA20 for a minute. It has a unit price of $180,000. It is a great little two-seater, very good safety record, economical, good performance, and popular as a trainer. Great little plane but who's got 180k? Can it be homebuilt for a lot cheaper? The consensus here seems to be no. But to help me understand why, as someone who is neither an engineer or an experienced composite builder, where does that heavy price tag come from? Some here are saying it is from the expensive tools. But for a one-off couldn't the plans creator obtain or make detailed 3d computer design files, as Cheapracer has stated, that would basically allow the builder to give that file to a CNC service and cut a mold out of something as cheap as wood? Or, if that is not quality enough, cut a plug with the CNC and make a fiberglass mold out of it. A bit more time consuming perhaps but can't be very expensive. What about the heavy price of engine, prop, and instruments. Well, the DA20 uses the Continental 240- 125HP with fixed pitch prop. Not terribly expensive, but perhaps you can find an experimental engine that fits the requirements, and reliable, for a bit less. And it probably has some higher end instruments, but one could always do the bargain hunting for those great Barnstormer deals and get some pretty good used instrument for a lot cheaper. So, again, help me understand why this plans built plane would be so expensive?

 

[autoreply]

Engineers who have designed planes like the KR2, or the Long EZ, or whatever, basically start from a clean slate and have to design everything from the bottom up using all their engineering knowledge and tools. And yet they are never really going to know what that plane is going to do until they have actually built it and flown it. It seems like it would be a lot easier to take an existing design, given everything known about the plane, including its flight performance and characteristics, and using engineering tools and composite design techniques to create plans that would yield a homebuilt airplane that was pretty much identical to a Cirrus. I'm not suggesting that a designer should do this in lieu of creating an original design, if they are passionate about designing airplanes then they would obviously want to create there own original airplane. I was just wondering if someone with engineering and composite skills could fairly straightforwardly create plans for a replica of a quality airplane such as a Cirrus or Diamond, because I think from a homebuilders standpoint this would be useful.

In the end, designing from scratch is often as quick, or quicker (and easier) as adapting an existing design.

Copying the whole exterior of a good plane gives you a safe baseline. Yet, simply having your tail volume in the right ballpark will too.

No matter what, you still have to work your way through all loadcases (a few hundred), since that data is not publicly available.

Why would it be useful? Let's switch our example plane to the Diamond DA20 for a minute. It has a unit price of $180,000. It is a great little two-seater, very good safety record, economical, good performance, and popular as a trainer. Great little plane but who's got 180k? Can it be homebuilt for a lot cheaper? The consensus here seems to be no. But to help me understand why, as someone who is neither an engineer or an experienced composite builder, where does that heavy price tag come from? Some here are saying it is from the expensive tools. But for a one-off couldn't the plans creator obtain or make detailed 3d computer design files, as Cheapracer has stated, that would basically allow the builder to give that file to a CNC service and cut a mold out of something as cheap as wood? Or, if that is not quality enough, cut a plug with the CNC and make a fiberglass mold out of it. A bit more time consuming perhaps but can't be very expensive. What about the heavy price of engine, prop, and instruments. Well, the DA20 uses the Continental 240- 125HP with fixed pitch prop. Not terribly expensive, but perhaps you can find an experimental engine that fits the requirements, and reliable, for a bit less. And it probably has some higher end instruments, but one could always do the bargain hunting for those great Barnstormer deals and get some pretty good used instrument for a lot cheaper. So, again, help me understand why this plans built plane would be so expensive?

It will be more expensive, if you use paid-for hours.

For such a production plane, figure on:

1. 100 hours labor for the main structure (all parts visible from outside)
2. 300 hours for the small parts (controls, hardpoints, etc
3. 100 hours finishing
4. 150-250 hours for installation of systems (engine, instruments, lights etc)

That's 650-7500 hours in labor. If you include actual wages (20-25 euro's/hour before taxes) and look at productivity and overhead cost for production (workplace, tools, hours spent doing other stuff), real cost is on the order of 80 euro's/hour. That's 52-60K euro's.

Direct cost:

1. Composites, resins and core: 5K
2. Parts and hard points 5K
3. Engine+prop; 20K euro's
4. Instruments etc: 20K euro's

So that's a total cost price of 102-110K in euro's. The tiny difference to the sales price (150K euro's) is what covers logistics (shipping to the US), marketing, amortaged certification and tooling cost, you mention it. No wonder few airplane manufacturers are barely profitable.


Now way you're not at least doubling those hours if you're building an amateur kit.

 

[Dana]

CNC cutting anything as large is a mold plug for an aircraft is not going to be cheap. Then, I'm guessing the production parts are cured in a [again large] autoclave, not something you find on every corner. An amateur, presumably, would have to use different types of composites and resins, so the design is already diverging. And details that make sense for series production don't make sense for a one-off.

That class of aircraft doesn't particularly interest me, I lean more toward low and slow and open cockpit, but if you look you can probably find an existing homebuilt design, kit or plans, that's comparable just about any small production aircraft with equal or better performance, that is suited to homebuilt techniques. Compare, for example, the Diamond with the RV line of kits.

Dana

 

[wsimpso1]

Tooling is big and either labor intensive to build or expensive to pay for. And there is a lot of it. Each side of each wing and the entire ouside of the fuselage has to be tooled in some manner. Conventionally, you would have a plug, then molds. CNC allows us to go direct to molds if you draft the entire airplane and its parts, with joints, etc, but it costly.

Yes, as a homebuilt, there are some tooling reductions specific to homebuilding. In many cases, hotwired foam for tail and control surfaces can be vacuum bagged and actually save weight over molded parts. Many flat or gently curved surfaces can be built on a table. Internal parts can be cut from large flat panels.

The big problem with re-engineering is you are doing more work than to design from scratch to suit you.

Billski

 

[TFF]

Your labor is not free. You might be doing it at no cost, but you you are either putting something else off or could be making money to offset how much work needs to be done. Lidar will give you a close copy outside, and you could probably get that done for about $5000 including the data crunching. CNC makes it low effort on your part but when you need a 1/2 to 1 million dollar mchine to cut the size parts you need, it will not be cheap as they need to make their mioney. Also with copying, you dont know what was learned on how it is assymbled; the lessons learned will be hard to come up with without making some of the same mistakes already made by the original designers. Mst RV-10s are professionally made; skip the legal aspect because they just are. The concept behind a RV10 is you can build a Curris like airplane for $100,000 yourself, or pay someone to build it and it cost you $200,000. that is a $200,000 savings on buying a SR22. At that level of homebuilt is all about big money. A guy up the road puts together Lancair PIVs; when they leave they cost up to a million. You will not do it for $20,000; if you were awesom at composites, you might do an airframe for $50,000. If you think there is a bunch of proffit margine at Lancair or Glasair you are wrong; they pretty much opperate at even money. With all the effort it would take why pick a POS like a DA20. It will take equal effort to do all this small or big so it might as well be a Cirrus or Corvallis.

 

[cheapracer]

[*]Composites, resins and core: 5K
[*]Parts and hard points 5K

Yup, and what happens from that point onwards is entirely up to the builder, their skillz and resourceful'ness.

CNC cutting anything as large is a mold plug for an aircraft is not going to be cheap.

So don't do it.

I have already demonstrated above but just one way to achieve a plug/mold. There are many others. Commercial retail CNC mold cutting is only recent.

Then, I'm guessing the production parts are cured in a [again large] autoclave, not something you find on every corner.

So don't do it.

I have been 'glassing on and off for 40 years, never used an autoclave. Never vac bagged either. I've done just fine.

I'm not the World's greatest 'glasser and I don't pretend I can whip up anything serious in terms of structural stuff, not something like a wing anyway, but hundreds of professional fiberglass guys who manage to produce superb results don't use autoclaves or vac bagging either - including some plane builders. ... and, NO, I am not saying those processes don't have their rightful place, they do of course.


Let's start finding ways people "can do" stuff, it might even appeal to "Homebuilders".

 

[PTAirco]

Yup, and what happens from that point onwards is entirely up to the builder, their skillz and resourceful'ness.

Exactly. If you want a DA20-type of airplane, figure out the cost of materials, engine, prop, and instruments and some plugs and molds and that is the cost of your airplane. The rest is your time and labor and resourcefulness. Your time is free despite what others tell you - yes, you could get another job and work on the weekends (can you...? Not an option for many of us anyway) instead and let somebpody else do the work; that might make sense if you're a lawyer or somebody who makes a lot more than the person you'd hire to do the work. And just how many qualified "composite airplane builder/designer/engineers" advertise around your area? Who charge less than you could make an hour? Most of us are in this homebuilt thing because we have more time than money.

Making a plug for the fuselage isn't rocket science - you don't need 3D scanners and CNC machinery; you need MDF, foam and Bondo and paint and elbow grease. No molds are necessary for wing/tail surfaces. I had a friend who could make plugs and molds at a the drop of a hat, he never built any other way and couldn't understand why anyone would. There is a learning curve, but it's not hard especially with help from others, like on these forums.

 

[autoreply]

Most of the remarks about autoclaves and prepregs are not applicable. Most GA aircraft manufacturers use hand-lamination, vacuum-bagging and a temper chamber. As such, save the molds, their processes can 1:1 be used by homebuilders.

To this date; I've not yet seen a composite plane that is really composite in terms of it's structural layout as well. Meaning highly integrated parts with many functions to enormously reduce build time (or cost, for a production plane). Of my quick summation in post 10, points 2 and 4 can easily cut by 2/3rd by designing for quick production in the first place. Unfortunately, that's rarely how designs are designed from the start. Once built, you're not going to re-design your design from the ground up...

 

[Aesquire]

The following is pure opinion.

There are different ways to "copy" an airplane.... or other complex object.

1. Copy in detail. This is how Antonov copied the B-29. Also how some warbirds are restored. It makes sense if you are pushing your company or nation to a new level of tech. The Antonov example is in spirit close to me wanting to build a Ferrari in my garage, after building choppers... they were under heavy pressure to pull Russia into a new tech level. The B-29 was a huge investment for America. A quantum jump in performance. (And a very flawed one put into production before serious bugs were worked out....I won't even get into the engines.) Likewise if you really want a P-51 & there just are not any on the market. ......

But it makes no sense if you can buy one. Warbirds are a very limited item. You can order 500 current production planes ( ok I can't, & you probably can't either, but a rich guy can )


2. You want a plane that looks like a P-51. Can't afford the real thing. Multiple folk have done that. With varying degrees of success. . Some replica WW1 planes are copies of the outside from drawings & photos, with all new structure & airfoil. ( because the original models have serious issues. )

3. You are inspired by a design and build one of the same configuration. That's a huge part of the last century of aviation.
Not really copying, there are some arrangements of bits that have proven to be practical.

As a buddy who is expert in 14th century armor says "they did it that way because it works"

NASA borrowed King Henry the 8th's foot armor to figure out how to articulate space suits.

 

[Nims11]

1. 100 hours labor for the main structure (all parts visible from outside)
2. 300 hours for the small parts (controls, hardpoints, etc
3. 100 hours finishing
4. 150-250 hours for installation of systems (engine, instruments, lights etc)

That's 650-7500 hours in labor.

Whoa, who said anything about labor costs? I'm talking about building a plane for personal use not starting a company selling replicas. I understand why it costs Cirrus that much money to build, but you are supporting the argument that it does not cost the homebuilder that much because there are no labor costs.

Your labor is not free. You might be doing it at no cost, but you you are either putting something else off or could be making money...

I thought this was homebuiltairplanes.com, where people come who are interested in building an RV or Long EZ because they want to build their own plane to fly, but also for the enjoyment, satisfaction, and experience of the build itself, and not just come here to talk about how difficult it is to build anything, in theory.

Direct cost:

1. Composites, resins and core: 5K
2. Parts and hard points 5K
3. Engine+prop; 20K euro's
4. Instruments etc: 20K euro's

This is what I'm trying to say, 20% of your entire material costs are the airframe, which is the heart of the plane, the design that gives it the flight characteristics you want. The other 80% is where some resourcefulness can save you a bunch of money. You don't have to buy a brand new engine right off, maybe you find a half-time Continental and prop online for a great deal and pay only 10k. You still have several years of flying life out of it before TBO, and in the mean time you can look for that auto conversion that just might work, or something else comes along that fits the bill. And same for the instruments. You can save a bunch of money finding used deals on instruments without buying the new Garmin GPS. It allows you to build and fly right now with the limited funds you have while upgrading your panel over time.

So, I've just reduced your figure of 110K to 30K, realistically.

 

[TFF]

"I thought this was homebuiltairplanes.com, where people come who are interested in building an RV or Long EZ because they want to build their own plane to fly, but also for the enjoyment, satisfaction, and experience of the build itself, and not just come here to talk about how difficult it is to build anything, in theory."

It is; still does not mean you time is free. You have to give it up and it is never recoverable, so it is very valuable. Make sure it gets you where want to go.

 

[Nims11]

CNC cutting anything as large is a mold plug for an aircraft is not going to be cheap.

There are people who sell CNC cut foam core for the long ez; wing, winglets, canard, and elevator, all for $2k. That is not that expensive. And they are probably using a 4x8 CNC table which you can build yourself from existing plans for about the same 2k. Maybe you need a bigger one that that for building plugs?

But I guess others are saying that CNC is not even necessary so I'll drop the CNC idea for now. Although it might be nice to have, not only saving time on your airplane build but having around for other projects.

In the end, designing from scratch is often as quick, or quicker (and easier)...

The big problem with re-engineering is you are doing more work than to design from scratch to suit you.

OK, well this is good news, and gives me inspiration. But it also makes me wonder why there are not more molded composite plans built designs out there?

And just to be clear, I'm not talking about building an exact copy of an existing airplane. Even designing from scratch you would surely use what is know about airplanes built and flown previously. What I'm saying is taking a Diamond, for example, and using all that knowledge about it, which can be obtained easily, and designing your own build process that gives you an airplane that looks and acts very much like a Diamond. Surely that has to be if not easier than building from scratch, at least more predictable in what you will get in the outcome.