When you click on links to various merchants on this site and make a purchase, this can result in this site earning a commission. Affiliate programs and affiliations include, but are not limited to, the eBay Partner Network.
Here are some detailed photos of a tower failure due to an on track crash. I don't know the crash details but this was a fully stripped down 981 car with a cage so I'm sure this was due to a racing incident. There was a ton of other body damage to the chassis that I didn't document.
All of the normal accessories that block the view of the tower structure have been stripped out.
Hopefully this will help the engineers on here further analyze the cause and potential solutions for this issue.
Pistol Pete thank you very much for posting these close up & detailed images of a typical failure. By noting that not a single bolt hole on the tower sides (there are 3) were compromised or cracked in your photos, I'm reassured that further securing the motorsport plates (with welded flanges) to these lower side of tower bolt holes as I've done & mentioned in my last post are a great way to reinforce this inherent weakness in the tower top.
I've been wondering the same. My leading hypotheses are:
1. Not enough on track and not enough getting decent miles (garage queens)
2. Longer front travel. The 991 front end has room to accommodate the lift kit, maybe that also means they have longer shock travel or a different front setup.
3. Maybe slightly different weight distribution of the mid-engine 981 puts more force into the front end.
#2 can be validated by comparing the front strut static position between GT3 and GT4. All things being as equal as much as possible, e.g., ride heights to min/max spec (choose one); similar options; and same amount of fuel. This can be done like motorcycle front fork compression tracking, with a ziptie on the rod. We have to remember that the GT4 was a parts bin build like the BMW 1M. Porsche may have certified the design by similarity and not run the whole gamut of tests.
#3 is a definite contributor. This can be validated by contrasting shock compression analysis on a given track. Some of our vendors pushing aero products have done this work already.
Thanks for the photos Pete. Unfortunately it is hard to make any analysis based on those alone. Inspecting a structural failure in an attempt to determine the root cause really requires high levels of magnification.
As I stated earlier, I am not a fatigue expert. I only studied it in engineering school many years ago. But for the benefit if this concern, in my opinion, there could be two failure modes
1. Fast Fracture: In this case, the design itself does not have enough margin for the load applied. The failure mode is catastrophic and occurs during one large load such as a track curb or road pothole. The structure of the body member is just not strong enough for the extreme conditions. This may be aggravated by a variable material condition like porosity in the casting or material defect.
2. Fatigue Failure: In this case the design may not have enough margin again but the way it fails is completely different. Repeated lower amplitude cyclic loading creates micro-fractures that grow over time. Eventually they become significant in size and when the structure is exposed to a high but "normal" load such as a track curb or hard braking, it is now not strong enough and fails. This could also be aggravated by porosity in the casting or material defect.
For condition 1, we would continue to just randomly see failures into the future. Condition 2 is more concerning because, if it is occurring, then the failure mode should increase over time.
Fatigue failures do leave specific visual cue of their existence. I found a few images on the internet showing what I describe. The failed material in the area of fracture will typically show two regions. One is where the crack slowly propagated. It will have different surface characteristics and may show signs of corrosion/oxidation since it was exposed to the atmosphere. The second is where the sudden failure occurred after it was weakened from the crack. The material fails differently in this region and it can be obvious visually.
The black area is the oxidized fatigue crack in a bicycle crank arm.
Added section for RHD versus LHD vehicles. I don't know what structural differences there are between the strut towers in LHD and RHD cars, but will try and keep the data as clean as possible.
Former BMW offered his information...thank you! Most of the other data points were obtained by reviewing this thread. Even though members have posted their information, I will keep their usernames private on the database unless given permission to make them public.
Added section for RHD versus LHD vehicles. I don't know what structural differences there are between the strut towers in LHD and RHD cars, but will try and keep the data as clean as possible.
Former BMW offered his information...thank you! Most of the other data points were obtained by reviewing this thread. Even though members have posted their information, I will keep their usernames private on the database unless given permission to make them public.
I think the driver side of the car is only useful in possibly indicating what side of the road the car is driven on. Maybe more important to indicate just that, what side of the road the car has primarily been driven on, for street driven cars. Probably makes no difference for cars that have seen almost all track use. Sorry for not thinking it through more clearly before posting.
This online PET shows the same part numbers for the shock tower parts regardless of USA or Europe application. I guess the USA or Europe selection indicates LHD and RHD respectively, but can't confirm that from information on the website.
For condition 1, we would continue to just randomly see failures into the future. Condition 2 is more concerning because, if it is occurring, then the failure mode should increase over time.
The database should capture miles (or km's) driven when the failure occurred. If it's Condition 2 then there will be an approximate correlation with miles/km's driven. Not exact, obviously, as road condition will play a big part. However, if the failure becomes more likely in high mileage cars then it points to condition 2.
The database should capture miles (or km's) driven when the failure occurred. If it's Condition 2 then there will be an approximate correlation with miles/km's driven. Not exact, obviously, as road condition will play a big part. However, if the failure becomes more likely in high mileage cars then it points to condition 2.
Itll be a mixed bag of both IMO. CrocGT4 on Instagram daily drives his and he’s well over 60K miles with no failure.
Worth comparing to the 992 piece? Photos by Pete Stout.
Originally Posted by stout
Cast front strut tower (dark gray), with Panamera-ish openings in unibody to either side that suggest how much dead weight cars have been carrying around for decades. Photo Pete Stout.
Itll be a mixed bag of both IMO. CrocGT4 on Instagram daily drives his and he’s well over 60K miles with no failure.
I agree, and I doubt we'll see a simple correlation between miles/kms driven and failures. Road conditions will play a big part. Yet, it's a useful piece of data as if we start to see more failures (and I hope we don't) then it's a piece of data we can use to help narrow down what's going on.
OK, some self interest on my part as I got my GT4 only 3-months ago with 17K miles on the clock already!
Great insight Sonorous... If it is in fact the latter, that is very frightening and discouraging, as it will eventually effect all cars, given enough time and use.
The market is ripe for a proper fix... some vendor must see an opportunity.
FWIW, I've heard of several failures on the clubsports, which all have the motorsport plates.
If you include the bracing strap picture, then yes;
Originally Posted by stout
Photo Pete Stout
I'm not an engineer, but I don't see how that brace has anything to do with the issue at hand. It's connecting the vertical wall - which is not the point of failure - to the frame rail. I suspect that strap is there for an entirely different reason. Perhaps reinforcing the connection of the tower to the frame rail. Also appears that there is a missing one that goes on the rear side.
In Pete Stout's photo below, those vertical reinforcement ribs on top will significantly reduce the high stresses around the edges of the top of the shock tower where we see the ruptures like in post #883. The 981/991 tower is smooth on top, and only a few small ribs on the bottom side. I would love to see the FEA results of both designs with the same GT car monoball shock/strut under the same load.
Interesting follow up to the Scottish, yellow, RHD strut failure that was posted on youtube. Apparently, both strut towers are being replaced as the other side deformed and is out of tolerance. He mentions "strengthening brackets" in the second link and I assume that means the motorsport plates, but will have to wait and see.
04-15-2019, 12:32 PM
