VansAirForceForums  
Home > VansAirForceForums

- POSTING RULES
- Donate yearly (please).
- Advertise in here!

- Today's Posts | Insert Pics


Go Back   VAF Forums > Model Specific > RV-8/8A
Register FAQ Members List Calendar Search Today's Posts Mark Forums Read

Reply
 
Thread Tools Search this Thread Display Modes
  #11  
Old 04-24-2009, 01:51 AM
scsmith scsmith is offline
 
Join Date: Jan 2008
Location: Ashland, OR
Posts: 1,884
Default bolts exceed mil-specs by quite a bit

Thanks Gil for printing the mil-spec chart for the bolts.

I'm fairly sure that the NAS bolts are 125 ksi, so that would say that they are required to achieve 10,300 lbs. from the chart.

Because of good bolt design, including rolled threads, 2-3 thread run-out at root, and radius under heads, most bolts FAR exceed the specs. So, although the rated strength for design purposes is 10,300, the actual strength is probably much higher. My GUESS is about 14,000 lb. When we get a chance, we'll break one and let you know.
__________________
Steve Smith
Aeronautical Engineer
RV-8 N825RV
IO-360 A1A
WW 200RV
"The Magic Carpet"
Hobbs 515 in 9 years (would have flown more this year if not for fire smoke)
also LS-6-15/18 sailplane
VAF donation Dec 2017
Reply With Quote
  #12  
Old 04-24-2009, 06:02 AM
Tomasz Tomasz is offline
 
Join Date: Dec 2007
Location: North Liberty, IA
Posts: 85
Default

Quote:
Originally Posted by chrispratt View Post
Steve:

So what conclusion am I to draw from this? Are you saying that the nuts Van's supplies are not adequate or does the test show they will do just fine?

Chris
IMHO as long as the nut's strenght is withing design limit I'd rather have weaker nut than bolt. In case it breaks it's much easier to just put new nut on and go home than get stuck with broken bolt. In many cases elements are chosen to act as fuses.
Reply With Quote
  #13  
Old 04-24-2009, 08:41 AM
terrykohler terrykohler is offline
 
Join Date: Jan 2005
Posts: 974
Default

Not sure what you're really trying to test here, but what you've really got is an apples and oranges situation. First, an ideal fastened joint is really one where you put the bolt or male fastener into a state of maximum elastic deformation. That is, by stretching the bolt to a point just before the plastic deformation point, you are creating max clamp force. Think of a bolt as a rubber band - no tension or stretch, no clamping force. That is why torque is used to "set" a joint. Even better, is to use torque angle, which really measures linear displacement of the nut as it moves around the bolt's helix. Sooo, if a joint does not require clamping force and the bolt merely serves as a clevice, than the only important number is bold shear strength. In a clamping joint, the important numbers are proof strength and tensile strength. These numbers are derived from cross sectional area and strength of material calculations and are properly tested by SAE and ASTM standards. Home-made tests don't mean much unless they simulate the actual application, and then only when the joint design and actual loading are properly understood. The only important number for the nut is it's proof strength, which is really a measure of the strength of the overall thread area. Finally, for those that see test numbers for locking or prevailing torque fasteners, the test numbers are typically:first on torque, first off torque, and fifth off torque. These are again done using standards from SAE or ASTM and are just to measure the performance of the lock-nothing more.
Terry, CFI
RV9A N323TP
Reply With Quote
  #14  
Old 04-24-2009, 09:04 AM
Bubblehead's Avatar
Bubblehead Bubblehead is offline
 
Join Date: Jul 2007
Location: Keller, TX
Posts: 1,527
Default

Great discussion. Another great question is do I want the bolt to stay in one piece while the tower rips out of the fuse, or would it be preferable to have the bolt fail and save the tower? I'm not sure anyone could design it that closely though because of variability of materials, construction, and loads.

As a 26 year engineer I learned long ago to define design intent before defining the design. I'm not sure what I would choose for design intent for this one.
__________________
RV-8 180 hp IO-360 N247TD with 10" SkyView!

VAF Donation Made 7/2017
"Cum omni alio deficiente, ludere mortuis."
(When all else fails, play dead.)
Reply With Quote
  #15  
Old 04-24-2009, 09:14 AM
az_gila's Avatar
az_gila az_gila is online now
 
Join Date: Jan 2005
Location: 57AZ - NW Tucson area
Posts: 9,621
Smile Perhaps you should...

...re-run the tests, but using the same type of bolt as called out in the plane?

The MIL-Handbook does say that bolt and nut fit specifications both need to be considered.

Quote:
Originally Posted by scsmith View Post
Thanks Gil for printing the mil-spec chart for the bolts.

I'm fairly sure that the NAS bolts are 125 ksi, so that would say that they are required to achieve 10,300 lbs. from the chart.

Because of good bolt design, including rolled threads, 2-3 thread run-out at root, and radius under heads, most bolts FAR exceed the specs. So, although the rated strength for design purposes is 10,300, the actual strength is probably much higher. My GUESS is about 14,000 lb. When we get a chance, we'll break one and let you know.
__________________
Gil Alexander
EAA Technical Counselor, Airframe Mechanic
Half completed RV-10 QB purchased
RV-6A N61GX - finally flying
Grumman Tiger N12GA - flying
La Cholla Airpark (57AZ) Tucson AZ
Reply With Quote
  #16  
Old 04-24-2009, 02:21 PM
Finley Atherton Finley Atherton is offline
 
Join Date: Nov 2005
Location: AUSTRALIA
Posts: 698
Default

Interesting results and full marks for doing it. I have not seen the landing gear bolt setup on the RV 8, but I am not sure how relevant this test is. In this test, given the large forces involved, there will be some stretching of the bolt and some bending of the bars. Both these actions will conspire to put most of the forces on the one side of the nut nearest where the force is being applied. In effect you now have a force trying to rip the nut apart in addition to trying to pull the bolt through the nut. In this situation, as you would expect, the nut with the smallest cross sectional area (the MS21042) would likely fail first. I have built two different aircraft and I can't think of any fixture that would put so much of the forces involved on the one side of the nut. The closest I can think of is where a stub axle is bolted to a leaf spring. In an extreme heavy landing there may be a tendency to load the lower bolt/nut in a similar manner, however this is really only likely if all the nuts on the stub axle were loose. As far as I am aware, the MS21042 develops the same tensile strength as the MS21045. It would be interesting to see a test where the nut is under tension without a significant force being applied to one side of the nut. I suspect the MS21042 would be up there with the MS21045.

Fin
9A

Last edited by Finley Atherton : 04-24-2009 at 02:33 PM.
Reply With Quote
  #17  
Old 04-24-2009, 03:09 PM
terrykohler terrykohler is offline
 
Join Date: Jan 2005
Posts: 974
Default

If the nut is torqued onto the bolt, the bolt is already in tension and the tensile load should be close to the proof strength. Unless the impact of landing exceeds this number, it doesn't matter - the forces are NOT additive. For example, if the strength of the bolt is 10,000 pounds and there is only one bolt per gear leg and the plane is landed on one gear, the impact has to be at least 6G's to overload the bolt. I suspect the gear leg will collapse before joint failure. In a properly designed joint, the nut is ALWAYS stronger than the bolt. That's how you generate elasticity. As a side note, the threads on the nuts are cut rather than rolled, yet they still exceed bolt strength. Many of these observations and tests previously cited are not relevant to the actual joint.
Terry
Reply With Quote
  #18  
Old 04-24-2009, 05:35 PM
scsmith scsmith is offline
 
Join Date: Jan 2008
Location: Ashland, OR
Posts: 1,884
Default

Fin: I assure you, the asymmetric loading onto the nut is MUCH MUCH smaller than you imagine. Three reasons.

First, the total elastic elongation of the bolt at the nut-failure load is .005 inches. About 1/3 of that was achieved through pre-load torque, and so only about .0035" of elongation is caused by the parts gaping.
Because of local yielding in the threads as the nut seats, and nut-embedding into the washer, the actual parts gap was probably closer to 0.007 - 0.010". So that is about 1/2 degree of rotation of the bar at the bolt.

Second, each half of the bolt is constrained in the close clearance hole in the bar, and can not rotate with respect to the bar - the 1/2 degree rotation produces bolt bending in the zone close to the slightly gapped joint, and where the bolt exits the bar to engage the nut, it exits very nearly perpendicular to the bar.

Third, local crushing of the washer under the nut will easily adjust and redistribute the load fairly well onto the nut.

For those familiar with the RV-8 attachment hardware, I guarantee you that the asymmetric loading of the nut in the actual assembly is MUCH MUCH BIGGER than the asymmetric loading in my test. This is because the saddle is rather thin ( average 1/4 inch) and there is an intended gap between the end blocks of the saddle and the wear plate. The springyness of the saddle makes it difficult to get a good preload in the bolt, and when the end block flexes, the bolt and nut see much more asymmetric contact than I have simulated. But in part, it was this asymmetric prying load that I wanted to approximate.

---------

I think many of you are expecting much more out of this test than was intended. I'm not trying to replicate all the loads that a landing gear assembly can be subjected to - in fact, that is impossible, and the design criteria used to establish design loads do not anticipate large unexpected loads during mishaps.

I AM trying to replicate one particular loading condition for which my engineering instinct suggests the hardware is a little bit under-designed. I have found that strength values for nuts is very hard to find. Generally the tacit assumption is that the nut is capable of developing the full strength of the bolt. Not the rated strength, or the proof strength, but the actual full bolt strength. I have found this assumption is sometimes wrong.

But big picture, if we are not hearing about nut failures, they must be ok. I would NOT recommend using a 21042 nut in this application. I will be operating from austere fields a lot, and I want more margin. I will use the NAS1804 nuts.

I'm trying hard not to be defensive here - with 29+yrs experience as an aerospace research engineer, I certainly DO NOT need a lecture on bolted joint design, behavior, and analysis. I'm resisting the temptation to give one!
Much of what has been stated here by others, especially by Terry Kohler, is either incorrect or only partially correct. Industry standards for bolt pre-load vary depending on application, but for general-purpose joints, the preload is typically selected to achieve 60-80% of bolt yield stress. The AN bolt torque charts are much less - closer to 35% of yield stress, and unless the added friction torque is accounted for when torquing lock nuts, do not even achieve that level of preload.
If any of you would like to discuss generally the topic of bolt torque and preloads, that is a separate topic that we can take off line.
__________________
Steve Smith
Aeronautical Engineer
RV-8 N825RV
IO-360 A1A
WW 200RV
"The Magic Carpet"
Hobbs 515 in 9 years (would have flown more this year if not for fire smoke)
also LS-6-15/18 sailplane
VAF donation Dec 2017
Reply With Quote
  #19  
Old 04-24-2009, 06:31 PM
scsmith scsmith is offline
 
Join Date: Jan 2008
Location: Ashland, OR
Posts: 1,884
Default 1200 lb drag load produces 10,000 lb at bolt!

For real: if you apply a 1200 lb load directly aft, at the axle, you will produce 10,000 lb tension in the front outboard gear attach bolt. That is over and above the 5200 lb of shear force that is shared between the two outboard attach bolts.

The distance from the rear edge contact point of the gear strut to the center of the forward attach bolt is 3.57". The vertical height from the axle to the gear attach is 30". ( assumed - I have not measured)
Moment equilibrium about the rear edge of the gear strut contact point gives:
1190 lb x 30" = 10,000 lb x 3.57"

I don't think it would take much of a pothole, or a big rock, on landing, to exceed that.
__________________
Steve Smith
Aeronautical Engineer
RV-8 N825RV
IO-360 A1A
WW 200RV
"The Magic Carpet"
Hobbs 515 in 9 years (would have flown more this year if not for fire smoke)
also LS-6-15/18 sailplane
VAF donation Dec 2017
Reply With Quote
  #20  
Old 04-24-2009, 06:56 PM
Bill Wightman's Avatar
Bill Wightman Bill Wightman is offline
 
Join Date: Aug 2005
Location: OKC, OK
Posts: 687
Default Possible mod to U-803 attach bracket

Steve, great effort and great writeup!

The thing I've looked a bit sideways at is the saddle (U-803 attach bracket) and how it constrains the gear leg from twisting due to drag loads applied at the wheel. That saddle thins down to around 1/8" or so where it spans the width of the gear leg... eek! So I'm a bit surprised to hear the 8's have had gear failures where the gear box was ripped loose. Maybe the gear box fails in torsion before the saddle bolts tear loose in tension.

Since the U-803 bracket is 4130 cold rolled, its soft and has fairly low tensile strength (90-95 ksi?), would it make sense to take those off and have them hardened up to around RC 30? That would increase the tensile strength of U-803 to around 140 ksi, or a 40% increase and it still wouldn't be too brittle to worry about fracturing.

edit: The distance on my 8 from the pavement to the center of the leg at the attach point is 32". I figured real quick that the bolts would reach 14000# with applied load of 1800# at the pavement. This looks to be in your ballpark.
__________________
“The important thing in aeroplanes is that they shall be speedy.”
- Baron Manfred von Richthofen


RV8 under construction
RV4 - Sold

United B777 FO, Chicago
Aerospace Engineer
F33A, RV8

Last edited by Bill Wightman : 04-24-2009 at 07:32 PM. Reason: added more info.
Reply With Quote
Reply


Thread Tools Search this Thread
Search this Thread:

Advanced Search
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

vB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Forum Jump


All times are GMT -6. The time now is 10:35 AM.


The VAFForums come to you courtesy Delta Romeo, LLC. By viewing and participating in them you agree to build your plane using standardized methods and practices and to fly it safely and in accordance with the laws governing the country you are located in.