87 replies to this topic

[Bootdog]

I know this is a bit of a hot topic at the moment, here's all the details of my toelink failure on track at Croft:

(I've been away for the past 2 weeks, so this is the first chance I've had to put this all together).

 

*** from the outset, the kit that failed was from elise-shop, not eliseparts as previously indicated by me and others ***

 

Firstly, what happened?

- a bolt failed on the rear left wheel toelink.

- I was halfway around a bend at the time.

- the car was instantly out of control, and swaying wildly.

- there was no warning, absolutely none.

- I went off the edge on the right hand side initially, then veered across the track and off the left side before stopping.

- the track was damp at the time, and the car wasn't under excessive corner load.

 

Video, in-car:

 

Video, from the car just behind:

 

Video, this is how much the rear wheels can turn when the toelink is disconnected:

 

Details of the car, the toelink kit, mileage etc.

- I bought the car in May 2011, it had been used for commuting, and had seen some track work, but is in good shape.

- Suspension was initially standard NA bilsteins, with uprated ARB installed 2 years ago, and exige bilsteins last year.

- I had the uprated toelink kit fitted in August 2011, mileage 90k miles.

- The kit was from elise-shop, described as "elise-shop toe link kit (Elise, Exige, Europa, 2-11, VX220)"

- In the interim time, I've done approx 12 trackdays, 2 activity days, and maybe 3k general use.

- It has been dry-stored for the past 2 winters, and I generally don't use it over Winter.

- Tyres have been Yokohama AD08 since 92k miles, with a new set of rears going on this year, at 101k miles.

- When it failed, the kit was almost 3 years old, and had done 13k miles, and say 15 trackdays equivalent.

 

[Bootdog]

Details of the failure:

- its the vertical bolt, with a socket head, it runs through the ball joint, and is secured by a nut on top of the upright.

- There are no signs of seizing on the bolt, but the nut was still stuck on top of the upright, so it wasn't loose or moving about.

- the old bolt is here, and shown in place, indicating where the failure occurred.

 

 

 

- the old bolt and nut is shown below, note the rusting, There is no sign of the (nylock) nut loosening.

 

- there are fatigue markings across the surface of the fracture.

 

- the old bolt is marked 8.8 on the head.

 

I managed to replace the bolt on the day with a bolt from the mango motorsport kit, I think it was a shock absorber bolt.

- the replacement bolt was an 8.8 bolt, and was torqued up to 45 Nm using an electronic torque wrench (thanks rally!).

- the replacement bolt has a hex nut head, not a hex socket head.

 

I checked it after going out, and everything was fine, so I tracked all afternoon on this bolt. I was putting much more load onto the car in the afternoon (dry track), and there were no issues.

 

Following the drive back to London, I took the wheel off again, to check for anything else that I might have missed on track.

 

I was shocked to find play in the same joint, with the new bolt in place. Its hard to describe, but I could move the joint a tiny amount (<1mm) and could get it to knock slightly. I checked the torque on the bolt, and it was pretty tight, perhaps lower than 45 NM, but not far off it, it was still reasonably tight. It was tricky to get the bolt out, and its clearly been bent:

Edited by Bootdog, 02 July 2014 - 10:51 PM.

[Bootdog]

How the failure occurred:

- it looks like the bolt bent initially, then the movement caused the bolt to be overloaded completely.

- a crack started in one of the threads, and opened a bit. The next hard corner opened it a bit more, etc.

- eventually, the cross section remaining (uncracked) failed in one go.

- there are marks on the old bolt indicating movement under load (cutting and shaving by the spacer piece).

 

A bit of engineering input:

- there are loads of threads floating about, with various theories about the bolts being sacrificial parts.

- the inner toe link was uprated to work in double-shear, which significantly reduces the stress in the bolt.

- this thread has some nice diagrams outlining the bending moment and single shear vs double shear.

http://www.lotustalk...-failure-33177/

 

- when in single shear, the bending moment which develops in the bolt acts in two directions on either side of the bolt

- it is stretching one side, and compressing the other (plus both sides are equally in tension due to 45 Nm torque)

- the fatigue mechanism is a combination of the static tension (45 Nm) plus the cyclic cornering load.

- fatigue failure is a massively complex area, particularly predicing when it should occur for a given static and dynamic load combination. 

 

For reference, I estimated the number of cycles as:

- 15 trackdays x 10 corners x 50 laps = 7500 cycles of a signficiant magnitude.

- this is a tiny number of cycles, typical lifetime ranges in fatigue calulations would be 1,000,000+

 

Loads of info here:, especially the links halfway down the page:

http://www.eng-tips.....cfm?qid=295351

 

Neat summary here:

http://www.materials.../CA-fatigue.htm

 

[Bootdog]

Root Cause :

- I think the combination of sticky tyres, and track usage caused it to be loaded up beyond its design.

- whether its considered "operation outside of the design limit", or "bad design" is debatable.

- then there is the question of which design : Vauxhall's? Lotus's? the aftermarket supplier's?

 

- clearly the toelink upgrade that puts the inner toelink into double shear is a big improvement on the original - it basically supports both ends of the bolt instead of 1 end.

- however, the upgraded outer toelink is still in "single" shear. If nothing else is changed, the point of failure is just moved from the inner bolt to the outer bolt at the wheel.

 

- in any case, the fact that the new bolt was slightly bent after 1 trackday indicates that the stresses were up near the yield stress of the metal (the point at which it "gives way" and becomes permanently bent.

- this is way beyond the safety factor for engineering design purposes (a factor of 50% or 2 is often used for design, where the design is optimised so that maximum stress anywhere in the assembly doesn't exceed 50% of the limit for the material).

- fatigue is generally considered in highly cyclic applications, from about 40% of yield strength.

- this application is clearly way beyond the "marginal" zone for fatigue, its probably more appropriate to ask why did it last so long before failure?

 

How to fix this?

- to me the obvious first route here is to use a stronger bolt, and replace it regularly (every 2 years or so).

- a higher strength bolt (10.9 grade) should carry more load without bending or failure due to fatigue.

- the stronger bolt will still be susceptible to fatigue, but it might move the failure point from 7,500 cycles to 750,000 cycles or 7,500,000 cycles.

- there is a case that the stronger bolts will be more susceptible to brittle fracture, but on balance I think the overall risk is lower.

 

- a bolt with fewer threads would also help, but I don't think it would solve the problem - there will be a thread somewhere along the bolt, which will be the weak point.

- there's also the AN bolt, which has pressed threads (not cut), which reduces the stress-raising effect of the threads.

http://spenceraircra...frame-bolt.html

 

There is precedent in this area:

- elise-shop offer a motorsport kit, which includes a 10.9 bolt set and upgraded ball joints:

http://www.elise-sho...0-p-502496.html

- in their own words:

"As the inspection intervals on race cars are much shorter than those on road cars, we also decided to move to a high tensile (10.9) bolt kit. 

To make life easy, we include one extra set of inner and outer bolts with each kit just in case you've bent something." 

- eliseparts also offer a toelink upgrade, which includes 12.9 bolts.

 

 

[Zoobeef]

Have you spoken to elise shop yet?

[Bootdog]

So ... Next Steps?

- I'm ordering 10.9 bolts and plan to change them regularly (say every 2 years).

- I'll also be checking for signs of bending (its unlikely that I'll be able to determine if fatigue is starting).

- I haven't opened the rear right toe link yet, so that might yield more info. I will update this thread with any new info.

 

Its worth pointing out that Yvo in elise-shop has been very helpful and open throughout, so I'll be keeping him posted on any further developments.

[Zoobeef]

Seen your answer. I jumped in too early. Majority of tracks are clockwise so would that be why that side failed first?

[Bootdog]

I guess so ... clockwise would cause LHS failures more than RHS.

I bent the replacement bolt (that I fitted at Croft) during the afternoon. That really surprised me.

 

[WrightStuff]

Interesting. Does anyone know what grade of bolt comes with the Spitfire kit ?

[RabnaKS]

I think you've gone to a lot of length to help find the weak link, and that's got to be helpful. Is it worth talking to the guys that do this kind of thing as a profession to give your theory some validity? But brilliant work, I enjoyed reading.

[Bootdog]

Yes, that would be useful ... and make sure to have a good look at the current toelink arrangement during the factory tour (I'd love to go but can't make it ...)

[mbes2]

Double toe link setup needed? My rear left also snapped. I plan to check my bolts as been around 2 years after upgrading Ghand crash ??

Edited by mbes2, 03 July 2014 - 05:32 AM.

[mbes2]

Double toe link setup needed? My rear left also snapped. I plan to check my bolts as been around 2 years after upgrading Ghand crash ??

 

My notes show 27/04/2012 - Uprated Toe links fitted - 49150 miles

[mbes2]

Whats this upgraded bracket fitted, seen here ?

 

[Kieran McC]

Double toe link setup needed? My rear left also snapped. I plan to check my bolts as been around 2 years after upgrading Ghand crash ??

May be a return to looking at a brace bar set up

Edited by kieranmcc, 03 July 2014 - 06:50 AM.

[mbes2]

This is from 2007, http://www.lotustalk...toe-link-42434/

 

also

 

http://www.lotustalk...e-links-136217/

 

lotus addressed the issue by doing 2 things (primarily for the inner joints)

 

1. offering a stabilizer bar (ultimately a factory add-on to the Exige in later years)

2. posting a bulletin on checking the torque of the joint bolts (apparently coming to the conclusion that the inner joints became loose thus causing them to shear over time from excess stress). if I recall, many dealers did offer to retorque the bolts as a recall service.

 

either way the current solutions for the rear toe link arms are primarily to address the inner joint from shearing (whether its BOE/Sector/ect they are all legit options)

 

but the stock outter joints aren't really a 'problem' spot (like the inners) & even after the unfortunate incident with 'Ying' I still wouldn't consider the outter joints an overall issue... but with that said if you address the inner joints with one of the aftermarket options, its certainly a good idea to check for wear on the outters as a piece of mind

Edited by mbes2, 03 July 2014 - 06:57 AM.

[mbes2]

VOSA ref [color=rgb(35,38,45);font-family:Arial, Helvetica, sans-serif;font-size:12px;background-color:rgb(242,242,242);]R2002/124[/color]

 

http://www.vosa.gov....&freeText=Blank

 

&

 

http://www.vosa.gov....&freeText=Blank

 

Recalls list

 

http://www.lotuscars...ustomer-service

 

 

A lotus service check list

 

http://wiki.seloc.org/a/Servicing

Edited by mbes2, 03 July 2014 - 07:06 AM.

[Exmantaa]

Reading all this I'm seriously considering to drill this corner for a slightly bigger dia. bolt and indeed maybe use an AN-type for piece of mind... 

Edited by Exmantaa, 03 July 2014 - 07:16 AM.

[peteslag]

Thank you Bootdog for taking the trouble to post all this information. I've just fitted the elise-shop kit so will keep a close eye on this thread to see what they do about it. I do about 5 track days a year so will definitely have to put in stronger bolts within the next 2 years or so. I think I'll also check the torque settings of the existing bolts periodically. 

 

Hopefully elise-shop will offer upgraded bolts for their existing kits.

[VXT Tim]

Interesting. Does anyone know what grade of bolt comes with the Spitfire kit ?

Just read through the spec sheet that Gaz at Spitfire sent me, it states 10.9 Chris.