Trailing Arm Bolt Engineering

or - More than you ever though possible to know about DeLorean Trailing Arm Bolts


Created: 11/13/01


Author/source: DML - Toby Peterson  tobyp(AT) katewwdb.com


This is a set of  multiple posts on this topic by Toby. He introduces himself in the first paragraph. Who would have thought this much could be studied about 2 bolts that are about 4 inches long (and happen to hold the rear suspension on the car)?

 

Reproduced here out of sheer admiration. You'll learn a lot about trailing arm bolts, and develop an appreciation for how engineers who design airplanes think about component failure. 

 

There are some VERY interesting people on the list.  ---das

 


 

11/5/01

Hello List -
I've got some information relating to trailing arm bolts that I think would be of interest.  First, let me give you a personal profile so that you know who I am.  My name is Toby Peterson, and I am, and have been, a Principal Structural Engineer at Boeing for almost 20 years.  My responsibilities include the engines and engine pylons for the entire 747-400 fleet.  I have owned my DeLorean, VIN 2248, since 1988. I have developed many connections within the aerospace industry and some of the best aerospace manufacturers in the world.  Now ... on with the story -

The trailing arm bolts (TA) have a great deal of work to do.  They react almost all engine torque and braking torque at the rear wheels, establish rear wheel alignment, and transmit all "thrust" from the drive wheels into the frame of the car.  They are a "critical load path" item with no significant redundancy.  If a bolt fails during certain driving scenarios, directional control could be lost, and the event could be non-recoverable.  The importance of the TA bolts has always been a concern of mine.

At a club-sponsored tech session last year, I did a complete inspection of the suspension components, as usual, and also re-torqued my TA bolts.  The drivers' side bolt took a very small torque, and then became free-spinning ... not a good thing.  As some other people went off in search of a replacement bolt, I removed both halves of the fractured bolt, and kept them for further inspection.  The other bolt was clearly bent, as well.  After the replacement, I took the bolts to a metallurgical lab for analysis of the fracture.  The bolt had cracked 80% through in slow crack growth, due to fatigue, with another 10% in fast growth.  The drive up to the session had been "spirited", and resulted in the last three crack striations.  The remaining 10% failed during the torqueing procedure.  Scanning electron microscope views of the fracture surface revealed that the crack had started at several small corrosion pits in the area of the first thread, and propagated through the bolt due to fatigue from bending stresses.  The material tested out as alloy steel with cadmium plating, and had a tensile strength of 136,000 psi.  That's about right for a bolt with a metric rating of 10.9.  After I explained where the bolt was installed, and what it did, the lab technician asked me a very simple question ... "Why did they use such a crappy bolt for this critical function?"  Good question.  The alloy steel is subject to rust and corrosion, the plating deteriorates over time and can be damaged during installation or use, and the material strength is not adequate to prevent bending in a single shear application under high loads.  As mentioned in other messages, the washers are showing signs of crushing and wear, which will reduce the preload on the bolts.  This will increase the induced bending stresses during driving, resulting in faster fatigue damage to the bolt.

At a subsequent tech session, we looked at the TA bolts in seven cars by completely removing the bolts and examining them visually.  Several were bent, and several others were corroded and rusty.  A couple were quite loose, while others needed to be pounded out with a hammer.  Only two cars had bolts in what I would call "good condition".  As I said earlier, I have been very concerned about this situation, and the apparent lack of understanding about this issue, as evidenced in other entries on the list.  The main issue with the TA bolts is not that they can cause a clunk ... the main issue is that a failed bolt can be catastrophic under some driving conditions.

I will post a second entry tomorrow with details about what I did to solve this problem for myself.  I will be asking for an idea of the level of interest in making my solution available to the rest of the DeLorean fleet.  Please consider what I have shared here, and be ready to give me some feedback when I share my solution with you.  'Til then...

 


11/7/01

 

During the tech sessions that I mentioned in my previous post, we were at a members' home with a mechanics pit in the garage.  The cars were sitting over the pit for the removal.  A jack was placed at the side being worked on, using the underbody jacking point, and the car is lifted until the tire just left the floor.  This allowed the wheel and trailing arm to be manipulated slightly to find the position where the bolt becomes unloaded.  It can then be tapped or pulled out. At most, you will see a slight shifting of the arm, but there should be no sudden movements to be concerned about.  Keep track of washers, shims, etc.  The bolts can be inspected or replaced with new bolts, placing the washers back into their original positions as the joint is reassembled.  As noted in other posts, there is some amount of wear on the washers, due to movement of components in the joint (a function of the "crappy bolt" bending under load).  I recommend turning the washers to provide a fresh surface against the sleeves in the arm and bushing.  We found that the bolt will slip back in easier if the car is lowered slowly until a small amount of weight is on the wheel.  This seems to get the holes in both the arm and the rubber bushing to line up better.  You snug the nut onto the bolt, lower the car to put full weight on the wheel, and then perform the final torque.


11/10/01

 

Hello List - This is Toby Peterson ... checking in.

I thought that I'd take a little time and give you some additional background behind the engineering considerations that go into
resolving issues like the trailing arm bolts (TAB).  I will try to be as brief as possible, and will also try to make it "value added" for everyone's learning.  The following terms need to be defined because I will use them a little later:  "Ftu" = allowable ultimate tensile stress;  "Fty" = allowable tensile stress at which the material starts to yield in tension;  "Fcy" = allowable compressive stress at which the material starts to yield in compression;  1,000 PSI = 1 KSI (reduces the number of zeros in an equation).  "Yielding" means that the material is beginning to deform and deflect under load.  When a material is stressed beyond the allowable yield values, it takes a permanent "set".  If it's a bolt, the bolt becomes bent.  If it's stressed beyond the ultimate tensile values, it breaks or ruptures.  Okay, are you still with me?  I think that this is important when discussing the various options, as well as the ultimate solution.

The original TAB are made from 4130 steel (probably), with Ftu = 125 KSI - 145 KSI.  My fractured bolt checked out at 136 KSI using the Rockwell hardness method.  For this strength range, Fty = 103 KSI, and Fcy = 113 KSI.  This material is also highly susceptible to corrosion, so it must be cadmium plated for protection.  The downside to cadmium plating is that it can be damaged by wear, and installation, and it's protection becomes compromised.  It's also sacrificial, which means that it dissipates over time.

The critical loading condition for the DeLorean TAB is bending.  We have a long, slender bolt in a single-shear joint.  We don't have significant tension loads applied during any driving scenarios, so the Ftu values don't really mean much.  The important numbers are Fty and Fcy, which define how resistant the bolts will be to bending stresses. We have all either seen or read about bent TAB's.  However, there are many people who have never experienced this problem (yet).  That means that the applied bending loads in our application are hovering in the range of the capabilities of the stock TAB.  Aggressive drivers have a very real concern that they will overload their TAB's, while the Sunday drivers' may never exceed the capabilities of their TAB's.  The way this bending phenomenon works is that the bending loads increase until the material in the bolt either meets the Fty or Fcy values.  Then, the bolt begins to yield in whichever manner is
critical for the material.  This increases the other stress dramatically, which causes the bolt to yield in both ways ... it will crush on the compression side, and stretch on the tension side.  If you exceed the maximum allowables, the bolt will be permanently bent.  The highest stresses will almost always be in the first few threads after the bolt shank.  If there are any corrosion pits or other damage such as galling of the threads due to installation of the nut, a crack may start at that point of maximum stress, and propagate through the thickness of the bolt.  Crack growth may be slow at first, because most of the bolt is still intact.  But, as the crack spreads, the stresses go up, and the crack speeds up.  It will eventually fail, just like mine did.

I have received a suggestion to use type 316 CRES for an alternate bolt material.  The numbers for 1/2 hard 316 are as follows:  Ftu = 141 KSI; Fty = 93 KSI; and Fcy = 61 KSI.  As you can see, for a given applied load, the value for Fcy is about 46% lower.  At a strength range of "full hard", it's still only Fcy = 83 KSI.  Not necessarily a good solution if bending is our primary concern.  I will say that 316 is very good for corrosion resistance, but ...

Is everybody still awake?  Okay, now for a glimpse of what I decided to do.  I selected the very best material that money could buy.  It's called Inconel 718.  This is a nickle-based super-alloy with the following numbers:  Ftu = 220 KSI; Fty = 200 KSI; and Fcy = 200 KSI.  Inconel 718 also has a very high fracture toughness, which means that it is very difficult to initiate and propagate a crack.  It's virtually corrosion proof, non-magnetic, and is used in the aerospace industry whenever a failure is absolutely not acceptable (engine mounts, landing gear, and wing attachments, to name a few).  I have developed a business relationship with the Vice President of Product Development at a world-class manufacturer of specialty fasteners for the aerospace industry.  This company is headquartered in Torrance, California, and supplies fasteners to Boeing, NASA, Airbus, and many others.  I selected a high-performance nut made from another super-alloy called A286, with an Ftu = 180 KSI.  The washers are made from hardened steel with a zinc dichromate finish.

I obtained a small number of bolts from this company, and have installed them on seven cars in our club.  The fit is perfect.  Due to the length of this post, stay tuned for what we need to do next to make these bolts available to concerned DeLorean owners everywhere. 


11/11/01
Hello List -

I will offer my conclusions on my studies of the TAB situation, without a lot of lecturing on my part.  All of the following are "in my humble opinion", and I invite discussion on any or all of it.

The key issue in the trailing arm installation is that the TAB's are bending under the loads applied while driving.  I don't believe that the issue is that the nuts are backing off, and allowing the joint to become loose.  The numbers suggest that the bolts are stretching and relaxing due to tensile yielding, because the numbers for that are somewhat lower than compressive yielding. The bolts are getting stretched slightly every time they are loaded up to the point of bending, and the little stretches, over time, will cause the bolt to get slightly longer (This is actually called 'creep').  This causes the bolt/nut to appear to come loose.  As the bolt stretches, the other components in the joint (washers, sleeves, etc) begin to move around as the bolt bends, resulting in wear at each point where the parts are pressed together.  This actually adds to the loosening of the joint.  All of you have either seen or heard of the wear and fretting on the washers, etc.  I better stop this ... I'm beginning to "go there again".

Bottom line - Yes, I have used science and engineering principles to design a bolt that will not bend or yield, at all, under the loads that I believe that we are seeing in this critical joint.  I have installed them first in my car, and then in several others.  I am absolutely convinced that I will never have any joint loosening or any more wear of any of the noted components in these cars.  I will never have to think about bolt rust or corrosion again.  I talked at length with the manufacturer, and he is willing to forego profit for these custom bolts.  He just needs to cover his costs of making them, so that the accounting department doesn't have a fit.  However, this level of quality is not cheap.  If you want the best, you have to pay for it.  But, you only pay once.  For a moderate-sized batch of bolts (200 pieces), with very good aerospace-quality NAS1805-7 self-locking nuts and hardened washers (for grip length adjustment when 2 or less alignment shims are installed), and including repacking and shipping to you, it's going to cost about $66 per car (2 bolts, 2 nuts, and 6 washers).  I am talking with Darryl Tinnerstet as the potential distributor for these.  My goal is not to profit from these personally.  My goal is to get rid of TAB's as a concern from a safety and reliability standpoint.  I need to get a good feel for whether there is a demand for these at that price point, so that Darryl and I can feel good about investing the money up-front in the first batch.  I've already "got mine", as do a handful of PNDC members.  The question is ... what do you want?  Peace of mind? 

Or ... not.  Please give me some feedback on this.

Toby Peterson, VIN 2248
Winged1
 


11/12/01

David (Teitelbaum)- You are quite correct in your assessment that the tubes and sleeves will collapse if the torque is increased significantly.  I installed the new bolts at a torque level of 50 ft-lbs, with a copper-based anti-seize compound applied to the shank and threads.  That is the same value as the "stock bolts" as I recall.  Forgive me, but I ran some numbers today on the preload in the bolts caused by the installation torque.  I wanted to see how much tensile stress we put on the bolts by torqueing to that level.  It revealed some interesting information ... without going into the calculations in detail, a torque of 50 ft-lbs on the TAB with grease on the threads and shank would create a tensile (tension) stress of about 116 KSI.  (Remember that 1 KSI = 1,000 PSI).  If the threads on the bolt and nut are perfectly clean and dry, the tensile stress value at 50 ft-lbs of torque is 48.3 KSI. Trust me on this ... it just works out that way.  The actual preload is probably somewhere in between those two extremes, but it varies depending on the cleanliness of the hardware.  The average of the two is about 82 KSI.  If you are brave enough to muscle through a previous post of mine, the maximum allowable tensile yield stress on the stock bolts is 103 KSI.  We are probably coming close to yielding, or stretching, the TAB every time it is torqued.  The variation could also explain why some people have a problem with their TAB's , and others do not.  My custom bolts have an allowable tensile yield stress of 200 KSI, so they would only get to about 1/2 of their capability at maximum torque.  Okay ... I'm done.  I just thought that you'd like to know.

 

11/13/01

Hello Group - I though that I'd "weigh in" on two points that were brought out here.  BTW - I am very happy to see some "considering" and "pondering" going on here ... it is important for everyone's learning to gather different viewpoints on specific issues.

For the first point, "brittle bolt breaking", as noted in David's first paragraph;  This IS a concern when the bolt is "pushed to it's limits".  Based on my previously noted calculations, we are at, or above, the yield limits for the current TAB.  The inevitable result is
bending to the point of yielding, fatigue damage, and ultimately, fracture due to fatigue cracking.  I also noted that the stresses are at about 1/2 the limit of my bolts.  You can't generate enough load in the system to reach the limit of 200 KSI yield strength in the Inconel material that I have selected.  The entire rear suspension (rubber bushings, washers, trailing arm assembly) will fail long before my bolts are stressed even close to their capability.

The second point that David brings up is the retorqueing issue for bolts.  He is absolutely correct in that, "When using bolts and nuts close to their yield point ...", retorqueing can result in stretching and yielding.  This allows joint loosening, which requires that you continue to retorque - until the bolt fails in tension.  As noted above, the current TAB is near it's yield stress point just from the
installation torque.  This is before any loads are applied to it from driving.  When my bolts are torqued, the stresses are far below the "elastic limit" (sorry for getting technical again!), and there will be no permanent (called plastic) stretching of the material.  The Inconel bolts can be torqued an infinite number of times, barring any damage to the threads, and never suffer from "plastic deformation".  Okay ... I'm done for now.  Thanks for your patience with me.
 

BTW - I have received indications of interest for 31 sets of bolts so far.  I will keep the list posted (so to speak) on whether this project will be economically feasible for myself and Specialty Automotive (www.delorean-parts.com), based on being able to move the majority of the first batch of 100 sets that I have discussed with the manufacturer.

Thank you, David, for your sound advice below on proper maintenance procedures that all should follow on every car.  When in doubt, read the instructions!  


--- David Teitelbaum.. wrote:

BTW the workshop manual calls for 55 ft/lbs see K:08:02-K:09:01. In some cases torque values are not created so much for the fastener as for the components you are trying to fasten together. I think in this case the limiting factor is the metal spacer tube in the pivot bushing. In most cases unless specifically called out torque values are for CLEAN, DRY threads. When using bolts and nuts close to their yield point or a critical fastener it is never a good idea to reuse (retorque) more than a minimum # of times. Every time you torque a bolt and nut you stretch them a little. After too many cycles you will just pull it apart or rip the threads out of the nut. On many of the newer cars where many bolts and nuts are tightened to high levels the manuals warn you not to reuse the fastener. (Another reason to refer to the manual for the specific car you work on!) BTW how often have you ever seen mechanics use a CALIBRATED (in the last 10 years at least) torque wrench on suspension fasteners outside of wheel lugs! There is much more variation in torque then you think!
David Teitelbaum
 vin 10757



Back to Technical Index


Home | Back Issues | Downloadable Files | Links | Subscribe
DeLorean FAQ Clubs and Events | DeLorean Mailing List FAQ


Copyright 2001 DMC-News
The legal fine print.
Comments, criticisms, questions:
webmaster(AT)dmcnews.com