Michelin problems at Indy 2005

[WhiteBlue]

more wanking

http://www.grandprix.com/ns/ns15111.html

A statement from Michelin

The French tyre company Michelin has issued a statement about what happened at the United States Grand Prix.

"Michelin has completed its investigations concerning the tyres used at Indianapolis and has communicated the results of these investigations to its partner teams," the statement said. "The tyres were not intrinsically flawed, but were insufficiently suited to the extreme racing conditions encountered through Turn 13 of the Indianapolis circuit this year. Given the evolutions concerning the cars' aerodynamics, the regulations which govern the sport and the nature of the track surfaces etc, Michelin carries out testing in the course of each season with a view to developing the tyres which are the most suited to each event. Two key elements must be known about the Indianapolis circuit: Turn 13, with its severe banking, is the only turn of its kind in a season of 19 races; Testing at Indianapolis was not possible.

"As a consequence, in order to define the specification of its tyres for Indianapolis, Michelin had to carry out simulation work based on the results of less severe testing at other venues and on estimations concerning the specific conditions likely to be met at Indianapolis in 2005. The Michelin investigations have revealed that the loads exerted on the rear left tyre through Turn 13 at Indianapolis were far superior to the highest estimations of Michelin's engineers. This year, the situation through this corner turned out to be altered by the extreme combination of the speed, lateral acceleration and additional dynamic load. The tyres which Michelin took were therefore insufficiently adapted to the extreme conditions of Turn 13 in 2005. This was a problem.

'On the other hand, investigations concerning the materials and construction employed for the tyres produced for Indianapolis have confirmed the absence of any anomaly. The tyres did not have an intrinsic flaw but they were not insufficiently suited to turn 13. Moreover, this analysis confirmed the pertinence of the tyre solutions specified for all the other circuits. In �retrospect, this analysis perfectly validates the pertinence of the precautionary measures requested by Michelin and its partner teams in the interests of driver safety and fully confirms that the addition of a chicane at the entrance of Turn 13, which would have guaranteed lower speeds through Turn 13, would have enabled spectators not to be deprived of a high class competition, while at the same time guaranteeing the safety of the drivers.

"As a consequence: Based on these investigations, Michelin has revised its simulation model for 'banked' corners such as Turn 13 at Indianapolis in view of the special effects caused by this corner; Michelin requests that it be possible in the future to undertake testing at Indianapolis before the Grand Prix; Michelin confirms that it will be present with safe, competitive tyres at the forthcoming Grand Prix races."

Michelin officially blames special dynamic effects of turn 13 to generate loads far superior to the maximum estimation of their engineers.

What do you think was this special dynamic effect if it wasn't resonance?

[strad]

Tell me WB...Is there any subject on which you are not more of an expert than the experts??

[WhiteBlue]

Pierre Dupasquier

http://www.motorsport-total.com/f1/news ... 20106.html

"Im Infield hätte es sicher keine Probleme gegeben, aber die Steilkurve kombiniert mit den neuen Rillen im Asphalt ließen die Belastungen nach oben schnellen. Konkret ging es um einen speziellen Punkt des Reifens, der zu schwach war. An der Stößelstange wurde eine Belastung von 1.400 Kilogramm gemessen. Gerechnet hätten wir mit maximal 900 Kilogramm."

Confirmtion for my estimate. Dupasquier confirms that the over load was 64%, and that one particular point in the side wall was affected. I also recall that the 2005 Michelin had a steel band in the shoulder. Perhaps that band introduced a spring factor which was unsuitable.

[Jersey Tom]

You are reading things that aren't there in your quote, WB. Pretty clearly reads that Michelin grossly underestimated tire loading (not just by "a few percentage points" as you commented earlier). As a consequence, you get overload failures. It's as simple as that.

No mention of resonance or crazy vibrations or any BS like that. FYI, "dynamic load" in this parlance refers to anything other than "static" load of the car sitting still in the pit lane. Just means loads when the tire is rotating. The so called "special effects" you latch onto are pretty clearly stated as additional vertical loading from downforce and the combined effect of lateral acceleration on a banked track.

But I do encourage you to try and find any case examples of actual "resonant" failures in tires - be they racing or consumer. I've never heard of or seen it. You're clinging to a bogus theory.

At the end of the day the story is pretty simple. As Michelin states, their simulated load profile was not representative of the (then) current track condition - and as a result had overload failures at the track.

[gixxer_drew]

Oval loads are crazy and I can see why they would be a real challenge. Gas pressure supports load and the tires I used/tested in ALMS were in the 27psi range hot IRL tires for the oval were ~45 IIRC. It is a major oversight, no doubt about that but don't underestimate oval loading. i could see how one thing or another could lead to having a bad expected load. I've seen that happen once before because someone didn't account for banking effect on the Y axis G sensor, when the other math channels didn't add up to the same value they though it was a strain gauge calibration problem. Their math checked out at every track before and after, it was years later that i figured out why their values were completely wrong there.

[WhiteBlue]

You are reading things that aren't there in your quote, WB. Pretty clearly reads that Michelin grossly underestimated tire loading (not just by "a few percentage points" as you commented earlier).

Plase go back and read exactly what i have written!

I have yet to read a single explanation why Michelin missed something like 30-50% of the relevant loads.

This was my estimation! It turns out that it was even 64%. You don't make such a gross miscalculation unless your model is wrong. Michelin missed a very important physical aspect and i have said it could have been resonance.

Dupasquier also mentioned soemthing that the French called "compage" in an interview with Ted Kravitz. It is described as bouncing. Please explain a bouncing that increased the load by 64%.

If we assume that four wheels were on average loaded with 85% of the critical left rear wheel we get a total peak downforce of 4.8 metric tons. This is a crazy amount of downforce for a 600 kg car. Michelin must have estimated 3 tons and they got something like 1.8 tons on top by their "bouncing". Someone has to do some explaining here.

[rjsa]

They fail to recognize that the Michelin problem was a physical problem caused by resonance. The side wall design was incited by an eigen frequency and nothing but a complete redesign would have cured that problem.

Michelin missed a very important physical aspect and i have said it could have been resonance.

Yeah, right.

[WhiteBlue]

Tell me WB...Is there any subject on which you are not more of an expert than the experts??

Because I care to read sources and have a big memory.

http://www.auto-motor-und-sport.de/form ... 57709.html

auto moptor sport said:

Die Michelin-Probleme in Indianapolis sind geklärt. Neben Vertikalkräften von 750 Kilogramm und Querbeschleunigungen bis zum Dreifachen der Erdbeschleunigung traten zusätzliche dynamische Kräfte auf. Die Reifen gerieten in hochfrequente laterale Schwingungen.

Diese waren so stark, dass die Michelin-Fahnder sie als sogenannte "stehende Wellen" sogar auf Fotos erkennen konnten. All zu sehr war man bei Michelin nicht darauf erpicht, rückhaltlose Aufklärung über die Ursache der Reifenschäden von Indianapolis zu betreiben. Erstens war das Debakel ohnehin nicht mehr zu verhindern, zweitens wollte man möglichst wenig über die Konstruktion der eigenen Gummis preisgeben.

Achillesferse Stahlband
Da Michelin im Gegensatz zu Bridgestone Reifen mit einer weichen Flanke, dafür aber mit einer schwereren Lauffläche verwendet, ist zur Stabilisierung der Schulter ein Stahlband eingearbeitet. Dieses löste sich unter den Schwingungen, die Schulter brach auf. Je härter die Hinterachse des betreffenden Autos abgestimmt war (wie bei Toyota), desto schlimmer die Probleme. Die weicher abgestimmten Williams war dagegen nicht betroffen.

Die eigenwillige Michelin-Konstruktion erklärt auch, warum die Michelin-Reifen so empfindlich bei Bremsplatten sind. Bei Unwuchten verformt sich der Reifen stärker als bei der Konkurrenz.

Translation:

The michelin problems from Indianapolis are analysed. Beside of vertical forces exceeding 750 kg and lateral accellerations up to three G there were additional dynamic forces. The tyres were exposed to high frequency lateral oscillations.

These oscillations were strong enough that the Michelin researchers could identify them as stationary waves on pictures. Michelin was not particularly interested to engage into unrestricted intelligence spreading regarding the root cause of the Indianapolis tyre failures. The debacel could not retrospectivly be undone and there was no interest to reveal more than necessary on the contruction of the rubbers.

Contrary to Bridgestone Michelin uses tyres with a soft shoulder and a heavy tread. For compensation a steel band is integrated to reinforce the shoulder. The bonding of the steel band failed under the oscillations and the shoulder cracked. The problem was more severely affecting cars with stiffer rear suspension (such as Toyota). The softer Williams in contrast were not hit.

The unusual Michelin design also explains why the Michelin tyre is so sensitive to flat spotting. The tyre deforms a lot more when unbalanced compared to the competition.

[Jersey Tom]

This was my estimation! It turns out that it was even 64%. You don't make such a gross miscalculation unless your model is wrong.

Vehicle model may have been OK, but the inputs to it (speeds, banking, driver line) totally off. By Michelin's own admission - which you quoted - that was the case. I suppose you could say that their track model then was grossly inaccurate, which I'd agree with.

This is a pretty well case closed deal if you ask me. Not a case of doing everything right and being totally blindsided by some magic effect that no other tire company had experienced before, because of some brilliantly different and unique design that Michelin was pioneering in 2005.

Michelin designed a tire to conditions not appropriate for Indianapolis. Plain and simple. (And/or had too much confidence in their sim outputs and did not allow for sufficient margin of safety overhead). People and companies make errors sometimes. That's the reality of the situation.

To clear up some other dangling comments: The "steel bands" - yes, Michelin use steel belts in some of their race tires. So do the tires on a minivan. No magic there. As for sidewall wrinkles - sure, I wouldn't be surprised if there was dramatic deflection and wrinkling there. That's what happens when you design a tire for way too low a vertical loading than it's going to see.

[FW17]

This was my estimation! It turns out that it was even 64%. You don't make such a gross miscalculation unless your model is wrong.

Vehicle model may have been OK, but the inputs to it (speeds, banking, driver line) totally off. By Michelin's own admission - which you quoted - that was the case. I suppose you could say that their track model then was grossly inaccurate, which I'd agree with.

This is a pretty well case closed deal if you ask me. Not a case of doing everything right and being totally blindsided by some magic effect that no other tire company had experienced before, because of some brilliantly different and unique design that Michelin was pioneering in 2005.

Michelin designed a tire to conditions not appropriate for Indianapolis. Plain and simple. (And/or had too much confidence in their sim outputs and did not allow for sufficient margin of safety overhead). People and companies make errors sometimes. That's the reality of the situation.

To clear up some other dangling comments: The "steel bands" - yes, Michelin use steel belts in some of their race tires. So do the tires on a minivan. No magic there. As for sidewall wrinkles - sure, I wouldn't be surprised if there was dramatic deflection and wrinkling there. That's what happens when you design a tire for way too low a vertical loading than it's going to see.

I do not agree to getting the model wrong story.

Michelin raced at Indy 5 times before 2005 so what had changed since the last time they were there?
The basic parameters for the model would have been same as what was used the previous year, the tyre was to last longer, but the loads would have been the same with the exception to downforce (which would not have made much difference at indy as it was a low downforce circuit)
So not a closed chapter

[bill shoe]

Oval loads are crazy and I can see why they would be a real challenge. Gas pressure supports load and the tires I used/tested in ALMS were in the 27psi range hot IRL tires for the oval were ~45 IIRC. It is a major oversight, no doubt about that but don't underestimate oval loading. i could see how one thing or another could lead to having a bad expected load. I've seen that happen once before because someone didn't account for banking effect on the Y axis G sensor, when the other math channels didn't add up to the same value they though it was a strain gauge calibration problem. Their math checked out at every track before and after, it was years later that i figured out why their values were completely wrong there.

This is actually a great insight. The banked Indy turn was the only banked turn of the year, and they may not have thought about the need to adjust the accel or load data.

[Jersey Tom]

I do not agree to getting the model wrong story.

Well it's effectively what Michelin reported, so call up Bidenbum and take it up with him. I'm also not entirely sure that Indy was completely unchanged between 2004 and 2005 but I don't like to speculate.

[WhiteBlue]

Jersey Tom,

the case is indeed closed because the AMuS article in conjunction with Dupaquier's figures completely confirms all elements of a resonance problem.

• massive dynamic overload of 56%
• high frequency lateral oscillation with a visible amplitude which wasn't supposed to be there by design
• visually identified wave form of the resonance as a stationary wave on the rotating tyre wall. You call it wrinkles.

The over load is caused by the wrinkles/lateral oscillation. They introduced additional stresses equal to at least 56% of the design loads. The tyres failed due to the root cause of these wrinkles, resonance.

The article even explains why Michelin wanted to keep the nature of their problem secret and used very unspecific descriptions. In 2005 their use of flat steel belts in F1 tyres was probably a less known design approach than today and they did not want to give any information about that to potential competitors.

I will rest my case here. Michelin will never confirm in writing what happened, but we have found all the relevant information to explain the failure mechanism.

[Jersey Tom]

The over load is caused by the wrinkles/lateral oscillation.

Other way around. Too much vertical load and the sidewall will wrinkle up. You can see this on test machines even when a tire has NO lateral force on it.

we have found all the relevant information to explain the failure mechanism.

We have. But you are still drawing the wrong conclusion. I also don't think there was anything unspecific about Michelin's description. Spelled out quite clearly to me.

[WhiteBlue]

So let us agree to disagree. There is no point to argue repetitively.