Quantcast

Load Test Question

HomeBuiltAirplanes.com

Help Support HomeBuiltAirplanes.com:

proppastie

Well-Known Member
Log Member
Joined
Feb 19, 2012
Messages
4,701
Location
NJ
I am at the 40% of my load test of the Aluminum Dragon. The total weight is 698 lb. on the LH wing for 5.33 Gs......The current weight loaded is 279 lb. which translate to 2.1 G..... I am getting serious oil canning on the buckling side (lower side) of the leading edge. The web and caps are not wrinkling or buckling. Is this normal, should I add span wise stiffeners to the first two bays.? I was taught the spar handles the bending. I may have reached my Peter Principal here.

When I remove the weight there is no permanent deformation (yet).

Shown sand bags in 10% increments. 20% to 100% Tractor jack to lessen tire deflection. Pictures of the leading edge deformation.

IMG_1153.JPG IMG_1155.JPGIMG_1157.JPG IMG_1158.JPGIMG_1160.JPG
 
Last edited:

BBerson

Light Plane Philosopher
HBA Supporter
Joined
Dec 16, 2007
Messages
13,759
Location
Port Townsend WA
Skin oil canning is ok if it returns to normal. I doubt you will get to 5.3 however.
I would add additional spar caps before it fails, if the cap starts to buckle.
The skin is buckling because the spar is under great strain. Stiffening the skin won't help the spar.
 

wsimpso1

Super Moderator
Staff member
Log Member
Joined
Oct 18, 2003
Messages
7,139
Location
Saline Michigan
A review of the fundamentals are in order here. At any one station along the wing, you have lift accumulating from the tip towards the station as shear, lift outboard accumulating as bending moment, and pitching moment of the foil accumulated from the tip to the station as torsion.

When you apply bending, it bends the entire cross section of the wing. The strain is directly proportional to how far you are above or below the neutral axis of the cross section. Everything sees the strain and the stress from this, with the strain growing linearly from zero at the neutral axis to max strain at the place on the cross section where you are furthest from the neutral axis. Likewise, the torsion accumulated twists the cross section about the centroid of the cross section, and the max strain from torsion is seen at the biggest radius from the centroid. Lesser trains are seen at smaller distances from the centroid. These two effects at every point around any cross section are superimposed.

Now the load goes where the stiffness is, so a beam sized to carry the load with a rather thin wing skin should still carry a large fraction of the bending, but the torsion is largely carried by the skins. And the skins are also seeing significant stresses. Our standard airplane design textbooks talk about crippling stresses when thin elements are under compression and have ways of calculating if those then elements will buckle.

I suggest that you do some research and then some calculations on your skins. Calculate the total stress state in the skins at a couple places and see if the crippling predictors say you skins should be buckling at this load, and is crippling is likely at full load. If yes, please do not get tempted to increase load and see what happens. If the calcs say you need more stiffness in the skins to get to full load, you might as well know it and get on it now, then resume test.

If you do the crippling checks and every thing checks out OK, but you are getting these wrinkles anyway, you might be OK. Just know that the wrinkled skin is now carrying a smaller fraction of the wing bending that it was, putting more strain and thus stress on the other parts that are still in shape. So, if your design passes the crippling check, I would look into how high the crippling stresses are in the various bays of the wing. Then think about stiffening all of the panels that calculate out as having stresses at similar fractions of the crippling strength before you put any more load on your test article.

Billski
 

BBerson

Light Plane Philosopher
HBA Supporter
Joined
Dec 16, 2007
Messages
13,759
Location
Port Townsend WA
In the photo it appears the skin buckles occur at the 2" spaced rivets but not at the 1" spaced.
The .016" skin tends to buckle at low stress at 2" space. It might just go back.
Keep an eye on the spar.
 

Heliano

Well-Known Member
Joined
Dec 24, 2015
Messages
107
Location
Campinas, SP, Brazil
Proppastie, it seems by looking at the pictures that you are loading the wing by hanging the ballast bags. Why don't you turn the wing upside down and simply lay the ballast bags on the wing? Aerodynamic loads are distributed loads by nature, not loads that act on a single point. You could do two things:
1 - Tilt the wing in pitch to simulate the L/D ratio. Normal loads (lift) always act simulataneously with longitudinal (drag) loads.
2 - Determine a priori where the airfoil center of pressure is and lay the bags at that percentage of the chord.
Static load tests carried out for the purpose of FAA certification (CFR 14, Part 23 or Part 25) are required to be applied for at least three seconds.
There are two requirements: at the rated maximum operational load factor there can not be any permanent deformation; at a load factor 50% above that one, there can not be structural collapse. Obviously carrying out the test to those limits is destructive.
 

proppastie

Well-Known Member
Log Member
Joined
Feb 19, 2012
Messages
4,701
Location
NJ
it seems by looking at the pictures that you are loading the wing by hanging the ballast bags
The distributed load is calculated such that the ribs have the proper portion of the load. The spar is at 27% so given the size of the sand bags that is close enough to the assumed center of lift, and final loading and I will have sand on top of the spar forward of he leading edge also. Tilting the wing was considered but I decided to forgo the complication in order to test the spar and aileron movement at load. You might notice the lack of traditional aircraft testing fixtures. I did park the tractor on a slight incline and tilting is still a possibility...but probably if it survives this test I will be ready for a nap. The caps are same top and bottom so to test the wing it is not necessary to turn over the wing upside down. I am trying to load to design limit load (max operation load factor) with no permanent deformation,...given that the design limit is 5.33 G I have decided to settle for that as a safe aircraft to fly.......This is a one-off aircraft as an exercise in learning aircraft stress analysis. Certainly to fully complete the educational process I should test to destruction but I am hopping to fly this aircraft,.....so there obviously are compromises in the process and project.


On my build log there are calculations for glider wing tip drag as regard the carry through fittings and rear spar fitting, these load numbers higher than the drag load.
 
Last edited:

BBerson

Light Plane Philosopher
HBA Supporter
Joined
Dec 16, 2007
Messages
13,759
Location
Port Townsend WA
I would test to 4g and then fly it. But never exceed 2g and fly only on calm days. That gives a safety factor of two. Then consider further testing options after some flights.
 

proppastie

Well-Known Member
Log Member
Joined
Feb 19, 2012
Messages
4,701
Location
NJ
I would test to 4g and then fly it. But never exceed 2g and fly only on calm days. That gives a safety factor of two. Then consider further testing options after some flights.
That is of course one possibility to test to say 3.8 G (normal category loading) and go flying. There are many good reasons to continue after the addition of calculated load stiffeners. To learn more..... do some calculations and prove out the calculations. Should at some point I desire to change the configuration such that the empty weight will take the aircraft out of the Part 103 category I will have these tests to convince a DER to license my aircraft. If I hit that really strong thermal I do not want to be wondering if my wing will fail up and block the BRS from working. I probably can think of more but you get the idea

I have made my calculations based on 12G ultimate load which some say is too high.....A load test to 5.33 does not seem unreasonable.
I will add I hope I will be able to calculate the panel buckling sized for a little above 5.33 and that will be determined by existing sizes of metal.
 

proppastie

Well-Known Member
Log Member
Joined
Feb 19, 2012
Messages
4,701
Location
NJ
I guess the conclusion is that this panel buckling is not normal and stiffeners are required if I am to obtain a reasonable strength of the Leading edge.
 

tdfsks

Well-Known Member
Joined
Aug 29, 2005
Messages
93
Buckling of skins and webs in a load test is entirely normal. However, in the case of your web, you have so many closely spaced stiffeners that there will be no tension field develop and that web is unlikely to buckle. It looks to me more like your spar caps (for want of a better word since it looks like just a bent flange) are buckling. How thick is that flange and attached section of web that is buckling ? Have you done any stress calcs if so what do you think the stresses are ? Looks very light to me for a cantilever wing ! You should be getting to ultimate loads (i.e. 5.33g x 1.5 = 8g for a sailplane) before there is a buckling failure.

Also you do need to think about your testing methods more closely. You are testing a negative case in those photos since the wing is upright. You need to turn the wing upside down to test the positive cases as these have the higher load factors and you want the test to be representative if you have buckling. As someone else suggested it is better to put the load on top of the wing although I understand that, if something fails, you want the load to be close to the ground and not fall on top of the wing.

You need to test two positive cases. In the first you should set the wing to approx 15-17 deg angle of attack so that the test load induces a forward component of shear in the drag plane of approx 20-25% of the lift load. The wing should be nose down with the wing upside down. In that case the loading will be closer to the leading edge - you need to calculate the center of pressure to know exactly where to put it. This case is VA/n1 or commonly called Positive High Angle of Attack (PHAA) and represents the condition with the aircraft at limit load at the maneuvering speed VA. In the second case you should test the condition where you have limit load at the design diving speed (VD) (i.e. n2/VD or Positive Low Angle of Attack (PLAA)). In this case set the wing at a lower angle of attack (say 5 deg) and apply the load further back on the chord since the center of pressure moves aft with increasing speed. How far aft depends on the pitching moment coefficient of your airfoil. In this case the shear load in the drag plane would normally act aft and if this is significant, you could also apply some drag loads. Both cases will be critical for different members and different failure modes.
 
Last edited:

proppastie

Well-Known Member
Log Member
Joined
Feb 19, 2012
Messages
4,701
Location
NJ
Ignorance, inexperience, enthusiasm certainly can lead to lots of learning experiences......

The Spar Caps are same thickness top and bottom, I believe that means it does not matter if the load is on a right side or upside down cantilever wing.

There certainly is not enough thickness of the flange I was relying on the spar caps to carry the load, see first sentence.

The gross weight is between 350-400 lb loading is for 400 lb gross.

I am currently reviewing my wing spar calculations.
 

TFF

Well-Known Member
Joined
Apr 28, 2010
Messages
13,644
Location
Memphis, TN
Can the skins be relatively “too strong” for the flexing ? Instead of following the flexing Of the spar, they are strong enough to resist. Not gathering up or stretching equal to the spar. Maximum strength per weight and resisting hangar rash may be not in balance.
 

BBerson

Light Plane Philosopher
HBA Supporter
Joined
Dec 16, 2007
Messages
13,759
Location
Port Townsend WA
The bottom skin has much less curvature than the top skin You can expect the almost flat skin to buckle under lighter load. Was D-cell skin buckling calculated at all?
If the wing was correctly positioned upside down for the load test, the upper skin buckling may not occur until higher g loads. I would probably continue as is, and not care about the skin buckling as long as it doesn't kink permanently.
 

tdfsks

Well-Known Member
Joined
Aug 29, 2005
Messages
93
The Spar Caps are same thickness top and bottom, I believe that means it does not matter if the load is on a right side or upside down cantilever wing.

I am currently reviewing my wing spar calculations.
Yes ... the spar maybe symmetric but what about the D box ? Your airfoil is not symmetric is it ? Also, it appears as though you have tubes to provide a drag truss (cannot see the detail). Is that symmetric ?

Skins are always going to buckle under bending loads and redistribute their load to stringers and spar caps. Go out in a Cherokee and pull 2g in a wind up turn and look out at the wing skins ...... :) You generally cannot rely on skins to carry any bending stress. Same for the shear web (you case might be an exception due to the extensive stiffening).

I suspect that the buckling mode we can see in your photo's is going to be hard to predict analytically although I would be checking the crippling stress of your spar flange as a guide. It will not be a high stress based on what I can see in the photos.

Post you calcs or send them to me privately. I would be happy to take a look.
 

proppastie

Well-Known Member
Log Member
Joined
Feb 19, 2012
Messages
4,701
Location
NJ
Post you calcs
sneaking up on the calculations....its a retirement project....so I try not to do too much to make it unpleasant........

Crippling of Z channel vs C channel.....conclusion is Z is higher moment I in the X direction than C and is a better choice.
1593610717210.png
 

proppastie

Well-Known Member
Log Member
Joined
Feb 19, 2012
Messages
4,701
Location
NJ
hanging the ballast bags
better to put the load on top of the wing
Sorry it takes a while to digest your help.....The hanging of the weights is from the top of the wing on the top of the ribs.....I add this as perhaps you thought I was hanging the bags from the bottom of the rib. I should have taken better pictures but I was a little upset at the time;
IMG_1165.JPG
 

BBerson

Light Plane Philosopher
HBA Supporter
Joined
Dec 16, 2007
Messages
13,759
Location
Port Townsend WA
The bottom wing skin only needs to be tested to about 2.6g. It only buckles in negative flight loads.
 

proppastie

Well-Known Member
Log Member
Joined
Feb 19, 2012
Messages
4,701
Location
NJ
The bottom wing skin only needs to be tested to about 2.6g. It only buckles in negative flight loads.
well we are pretending the bottom is the top with the symmetrical spar, even though the more highly curved top leading edge skin should take a little more crippling stress as was pointed out.
 

wsimpso1

Super Moderator
Staff member
Log Member
Joined
Oct 18, 2003
Messages
7,139
Location
Saline Michigan
Can the skins be relatively “too strong” for the flexing ? Instead of following the flexing Of the spar, they are strong enough to resist. Not gathering up or stretching equal to the spar. Maximum strength per weight and resisting hangar rash may be not in balance.
Tough to feature that in my thinking. Usually, you start with the assumption that all of the structure is tied together, contributes to bending and torsional strength, moves together under load, etc. Adjust sizes of parts until My/I is safe. Then check for buckling and crippling. If a cap or skin buckles or cripples, you can assume that the element is now carrying much less load. You go back through the calculation of neutral axis and I without that element, and check if the structure is still OK. If it is, then you can do the whole check of tension field stability, elastic stability, etc. If it is OK then, you might hazard a test. If doing the crippling check leads you down a path that says the whole thing will collapse, well, you have some more work to do before a test makes any sense.

I think what is being suggested is that the wing skin is stiff, the spar cap is more flexible and buckles first. Usually our skins are thinner and softer than the spar caps, and they buckle first. Usually spar caps are connected to the shear web and the ribs pretty stoutly and so is pretty resistant to crippling, etc. Skins that are thicker and more stout than the spar , well, I suppose it *could* happen, but it would take some mighty thin spar caps that are also lightly connected to the shear web and to make them less resistent to crippling than the skins. If you did have such a situation, rivets closer together connecting the skin to the caps and connecting the caps to the web would go a long ways to fixing it.

Billski
 

wsimpso1

Super Moderator
Staff member
Log Member
Joined
Oct 18, 2003
Messages
7,139
Location
Saline Michigan
I just went back through this thread.

First, this cambered wing form does tend to make a non-symmetric foil even before the skin in compression buckles. Much of the cross sectional area of the wing is below the centerline of an even symmetric spar, which moves the neutral axis and influences the I and y's you will use in calculating stresses. Then when you get to the threshold of buckling the skin in compression, the load in the skin that is buckling is all the higher the load in that skin will get, while the rest of the structure will continue to pick up moment. Neutral axis will shift away from the buckled skin, so your stress picture changes as the load goes up.

Having the buckling start this low - yes, I would go looking at crippling stress calcs and stiffeners, maybe some rivet lines can go to smaller spacing. My usual approach to early failures is to use thicker spar caps, which reduces bending strains. Remember that you need to double the load at which buckling commences. In this metal structure, I would look at bumping the spar cap thicknesses and look at adding skin stiffeners. The lighter one would get more attention from me. Give the wide gulf between load where buckling began and where you have to get, some spar stiffness and skin stiffeners might both be in order.

Billski
 
Top