Quantcast

Testing ?

HomeBuiltAirplanes.com

Help Support HomeBuiltAirplanes.com:

Pilot-34

Well-Known Member
Joined
Apr 7, 2020
Messages
583
Is it possible to do sufficient economical testing on a small plane
(Under 3000 lb )
To be reasonably sure it’s safe ?
To give you a idea of what I mean reasonable tests vary with the cost of failure.
I can deal with testing a wing rib to failure .
And I might handle a 80 %
chance of a wing spar failing in testing.
I think I could stand to destroy a wing while building wings. So a test with a 30% chance of destructive failure might be acceptable
But maybe not after the plane was completed. And certainly not if the test was likely to destroy the finished plane.
A 10% chance is about the max I’d be willing to take there.
Would it be possible to do that testing under $10,000? , $50,000?

In other words is it realistic to trade engineering for testing ?
 

Topaz

Super Moderator
Staff member
Log Member
Joined
Jul 29, 2005
Messages
14,178
Location
Orange County, California
Is it possible to do sufficient economical testing on a small plane
(Under 3000 lb )
To be reasonably sure it’s safe ?
...
"Static Load Testing" is the usual structural testing method for both large and light aircraft. A Google search will turn up quite a bit of information, as will a search here on HBA. That's a hint, by the way.

In other words is it realistic to trade engineering for testing ?
To a certain point, if the structure doesn't deviate very far from something already known, the answer can be a qualified "maybe." The problem comes in determining with how much load a given part of the structure should be tested. Determining the amount of testing load is not much less difficult, in terms of engineering, than basic structural engineering in the first place. So if you can do one (and you have to, in order for your testing to be valid), why not do the other?
 

Hot Wings

Grumpy Cynic
HBA Supporter
Log Member
Joined
Nov 14, 2009
Messages
7,263
Location
Rocky Mountains
The problem comes in determining with how much load
Thatone
This is the foundation of the whole structural design phase. Load testing a seat structure to X "Gs" is pretty simple. You know the pilot weight and can pile on sand bags as needed. But figuring how hard the air is going to be pushing back, and where, to make that same "G" load isn't as simple.

Someone that has built a lot of similar structures, and paid attention to where and how they failed, can sometimes come pretty close using TLAR methods. Those guys are few and far between in the aviation world.
 

wsimpso1

Super Moderator
Staff member
Log Member
Joined
Oct 18, 2003
Messages
7,119
Location
Saline Michigan
Hmm. As a 37 year experienced mechanical engineer who knows about all of the different load cases we have to design our airplanes to tolerate, I am hesitant indeed to say that you can test and say your airplane is safe. Here is why:

We have max positive g at stall AOA, Vne at whatever AOA gives us max + g, we have max structural cruise speed at what ever AOA gives us max +g, we have the mirrors of all these at max negative g. We have these cases at forwardmost CG, aftmost CG, and everywhere in between. Then we throw out the flaps and must be OK to the max flaps airspeed and whatever the max stated positive g is for the bird. This puts up a bunch of cases for the wings, tail planes, fuselage, and mounts for everything that goes inside the plane. We also have to stand a thumper of a landing without getting over 3 g and without bending or breaking anything.

The control surfaces all have rules for their structures based upon wing loading and design max g. Then there is the skin on the outside of the plane. Remember that the air inside is stationary and the air outside is moving some amount faster than the airplane is moving, which tends to inflate the airplane and further stress the skins and ribs and frames and longerons.

Then there is the 19g and 23 g crash pulses on the seats, restraints and other stuff in the cabin.

Then we get to the overload cases, where things can bend but not break. All of those case above have to be OK, but now we are at higher g for aero loads and interior loads, plus a higher drop height for the gear tests.

So, imagine trying to build enough test pieces and test airframes to do all of these cases, then repeat the tests after stuff fails and you beef stuff up. Not really a good scheme. Better scheme is to design the thing from the beginning to pass all those cases, figure out which cases are the worst cases for each piece of the puzzle, and then come up with a scheme that allows you to definitely say "if this piece passed this test, it is OK on the others!" This will significantly reduce the number of tests. Bumping you design a tiny bit to make sure that you do not bend stuff in one test, so you can use it in other tests is also a good idea.

Now, how do you do all that? Engineering. Design calcs, analysis, exemplar testing, and so on. You have to KNOW from accepted engineering methods that one case is worse than the others. And it is a really good idea to be right on it passing the tests. Then you can run one big test of the wing and show that it went above max operational load. Repeat on the horizontal tail, the control surfaces, the landing gear, the engine mounts, seat belt mounts, baggage containment, etc.

If you do all of this, and have done your homework on prevention of fatigue crack growth and fracture mechanics, then you can be reasonably sure that you will not break up the primary structure in flight. That just gets you past the primary structure. Then you have the skins, ribs, longerons, etc under those inflation loads I talked about earlier. After all of that, you still have control systems, fuel system, electrical system, any hydraulics, doors, windows, avionics, and so on.

So, just because you can put a measured set of loads on a wing and it is OK does not mean the airplane is safe...

Billski
 

Dana

Super Moderator
Staff member
Joined
Apr 3, 2007
Messages
9,364
Location
CT, USA
In most cases, engineering analysis is easier and more reliable than testing. Testing may have value where the engineering is uncertain, for example in the case of novel materials or structure, but to get meaningful results you still have to do the engineering to determine what loads must be applied where to adequately mimic the actual flight loads.
 

Dan Thomas

Well-Known Member
Joined
Sep 17, 2008
Messages
5,469
For a non-engineer it would be far smarter, far better, far less expensive and far less time-consuming to select a proven design and build it from kit or plans. None of us live forever and most of us have limited finances and time, so deciding that you're going to build some unique airplane "just because" could easily mean you never fly at all. Just building from plans can take years. None of this is as simple as some uninitiated folks think, and if the numbers haven't changed, when I joined EAA in 1972 the completion rate of homebuilt project starts was about 10%. Might have gone up some with kits, but kits also cost more money and there will be lots of unfinished kitplanes out there, too.
 

Wanttaja

Well-Known Member
Joined
Sep 15, 2013
Messages
1,559
Location
Seattle, WA
Is it possible to do sufficient economical testing on a small plane
(Under 3000 lb )
To be reasonably sure it’s safe ?
To give you a idea of what I mean reasonable tests vary with the cost of failure.
I can deal with testing a wing rib to failure .
And I might handle a 80 %
chance of a wing spar failing in testing.
I think I could stand to destroy a wing while building wings. So a test with a 30% chance of destructive failure might be acceptable
But maybe not after the plane was completed. And certainly not if the test was likely to destroy the finished plane.
A 10% chance is about the max I’d be willing to take there.
Would it be possible to do that testing under $10,000? , $50,000?

In other words is it realistic to trade engineering for testing ?
Depending on the size/complexity of the aircraft, in my opinion, it's a reasonable tradeoff.

In the US, homebuilt aircraft are licensed without the designer/builder/owner having to prove that it meets any sort of structural standard. It's not that way in other countries.

Over 40 years ago, a Fly Baby builder in Finland wanted to get his airplane licensed there. Finland is one of those countries were there is (or at least was, I don't know what it has now) no "shortcut" for getting approval of a homebuilt aircraft. All aircraft, including the man's Fly Baby, had to be tested to, essentially, show the plane met the structural and other requirements of FAA Part 23.

They did it with static testing....using sandbags to simulate the loading the plane would see in flight.
finnish_diagram.jpg
There are more than just this test, of course... similar loading tests were performed on the engine mount and other major components. You can find the complete test write-up (in English) at:


The original Finnish report is hand written, and does include photos of the test setups. However, the copy I have is so bad it's difficult to make out what is happening.

However, depending on the complexity, it may be difficult or impossible to devise a test that will adequately verify the aircraft. If you're still looking at an amphibian, for instance, there's a bunch of water loading aspects included in Part 23 that may be difficult to verify by static test. Of course, as a homebuilt, you are not required to provide any sort of verification, analysis or test.

Ron Wanttaja
 

BBerson

Light Plane Philosopher
HBA Supporter
Joined
Dec 16, 2007
Messages
13,733
Location
Port Townsend WA
Is it possible to do sufficient economical testing on a small plane
(Under 3000 lb
Just build a partial unfinished wing to test. For example, if sheet metal build the D-cell box structure and bolt it to something (I have a 5 ton truck). You can watch the structure buckle slowly and see how far you want to go. If it looks bad take it back and make a stronger spar.
It doesn't really cost much if it's your hobby and time only. Might take many years, depending on design.
I recommend a single seat first design.
 

ScaleBirdsScott

Well-Known Member
Joined
Feb 10, 2015
Messages
1,269
Location
Uncasville, CT
Start by copying something else. Then do a test. It shouldn't cost much at all to just build a wing or set of wings.

If you're doing something more novel, 10-50k can pay for some actual and vital engineering. Turns out even just having someone figure out what the loads are and how to apply them takes a fair bit of engineering math. But if you were, say, basically cloning a Kitfox type wing, with effort I think a non-engineer could pretty-much figure all of it out enough to make it work. As long as some research is done into it and there are no "I dunno why they made it like that, I'm gonna simplify it" moments.
 

Mad MAC

Well-Known Member
Joined
Dec 9, 2004
Messages
664
Location
Hamilton New Zealand
Having supported a type certified aircraft that was mostly justified by structural test, making changes to the orginal design down the track can be a right pain in the XXX.

But for a one off that isn't radical and most of the detail can be copied from other similar class aircraft, one should be able to get most of the way there with structural testing, just be prepared to build two sets of the most critical items (limit load testing is pretty hopeless at the best of times, and is of very limited value here). If you are adverse to the math best to farm out determining the loads and the critical load cases, although you would need to find a pragmatist engineer, its too easy these days to run all the load cases rather making a judgement as to the worst, which is what is needed if one is going to structurally test.

In the aforementioned aircraft, they structurally tested the wing spar alone & then proceeded to using it in two wings of the same plan but with different aerofoils ( a 6 series and a 23xxx).

There is a lot to gained structurally testing by careful technique, for one GA aircraft's test program, when testing the (metal) fuse, every time they got to the edge of failure they would back it off and add a small local reinforcement and repeating until the design load was reached.

The other thing to note is that ultimate load testing is likely to make the struture unsafe for flight without rework (stretched skins, rivets, permanent set, etc, etc).

The Light Aircraft Association have some interesting guidance
 

Dan Thomas

Well-Known Member
Joined
Sep 17, 2008
Messages
5,469
Depending on the size/complexity of the aircraft, in my opinion, it's a reasonable tradeoff.

In the US, homebuilt aircraft are licensed without the designer/builder/owner having to prove that it meets any sort of structural standard. It's not that way in other countries.

Over 40 years ago, a Fly Baby builder in Finland wanted to get his airplane licensed there. Finland is one of those countries were there is (or at least was, I don't know what it has now) no "shortcut" for getting approval of a homebuilt aircraft. All aircraft, including the man's Fly Baby, had to be tested to, essentially, show the plane met the structural and other requirements of FAA Part 23.

They did it with static testing....using sandbags to simulate the loading the plane would see in flight.
View attachment 95493
There are more than just this test, of course... similar loading tests were performed on the engine mount and other major components. You can find the complete test write-up (in English) at:


The original Finnish report is hand written, and does include photos of the test setups. However, the copy I have is so bad it's difficult to make out what is happening.

However, depending on the complexity, it may be difficult or impossible to devise a test that will adequately verify the aircraft. If you're still looking at an amphibian, for instance, there's a bunch of water loading aspects included in Part 23 that may be difficult to verify by static test. Of course, as a homebuilt, you are not required to provide any sort of verification, analysis or test.

Ron Wanttaja
In the '70s I bought a damaged Taylor Monoplane that had been poorly built. The right wing had bad wash-in and the airplane must have been a real scare in flight. It had maybe six hours on it. There were many other problems, so once I rebuilt the wings and forward fuelage I did a static test similar to your diagram. At the time I was working in a truck parts outfit and had a warehouse full of heavy brake drums and wheels. I supported the fuselage at the longerons at the cockpit, upside-down, and started loading the wings with brake drums. Got to just under 2G when I found that the left aft spar was cracking. I think it might have already been cracked, and the load revealed it as things got pulled. This was an airplane designed to +/-9G, so a 2G failure was a VERY bad sign. Gussets inside the fusleage were also popping off. The glue (Aerolite) had dried instead of curing, due to the original build in a cold workshop. The airplane became a cardboard box full of instruments and hardware. It wasn't worth taking it any further than the three years I had put into it already.
 
Top