A few people have asked me about shaker work. Thought I would post the link now that we are firmly in the season...

http://www.bimmerworldracing.com/data/images/P1010006.MOV

Cool stuff!

Prodded to do this today while our engineer was working on some things for me. Here is the text from him on his project that explains it some:

Shown is Seth Thomas’ BimmerWorld, BMW 325i on the shaker rig at the Penske Technology Center (formerly the Auto Research Center) in Mooresville, NC. The car is undergoing a “frequency sweep” in which its four corners are subjected to inputs of increasing frequency.

The test, conducted in early January of 2005, allowed BimmerWorld race engineer and technical advisor Dick Golembiewski, to better understand the car’s response in pitch and heave when subjected to those inputs. It also allowed the tem to optimize its Moton damper settings in order to improve mechanical grip.

“You have to keep the tires on the ground in order to maximize grip”, said Dick. “The shaker rig test allowed us to see how the car responded to certain inputs, so that we can reach the necessary compromise between mechanical grip and pitch oscillation.”

While on the rig, the Bimmerworld team also measured the stiction in the suspension system, and evaluated some different damper and structural hardware.

One wonders what a session like that would cost? I'm sure the information gathered was well worth it when setting up a full-effort racing venture.

Were all the BW cars strapped down... or just one with the findings carried over to the others at a later date?

Thanks for the video!

Technically, not really strapped down - given enough force, you would launch it in the air. That is the point - to calculate how the tires react with the surface with no external forces. This is a 7 post machine, but we didn't get into simulating downforce during the session with the video.

Right - it isn't cheap. The equipment is well over a million and to make money they have to charge...

No need to do all three cars. You are working to test theories and designs. Our cars are copies so everything transfers directly - that is important. But the information also transfers to other cars using similar parts and even the theories to cars with completely different setups.

Technically, not really strapped down..

yeah... poor choice of words on my part :(

Cool that you guys were able to use that equipment and benefit from the data.

yeah... poor choice of words on my part


Yeah - it is a wierd concept. When you test pitch, it is kind of amazing that the car doesn't go flying.

I'll have to watch the video again tonight.... but it looked like as the frequency picked up, the car's movement slowed down... which is kind of what one would expect... but it is strange to see it in an environment like that.

Very cool data points.

I assume the Spec tires don't change during the season. That could seriously change the damper settings.

Very interesting, I wish Dick could have come to Fsae, would have liked to have some chats with him.

Goodluck with the development, shaker rigs are sweet.

I assume the Spec tires don't change during the season. That could seriously change the damper settings.

Toyo's seem to stay pretty consistent from batch to batch. As far as I'm aware, the Toyo's World Challenge runs are no different from the Toyo's you can buy from Vilven or other distributors.

So Gvl, when is BFG and/or Michelin going to get back into the game and produce a R-comp to compete with Kumho and Hoosier? :)

So Gvl, when is BFG and/or Michelin going to get back into the game and produce a R-comp to compete with Kumho and Hoosier? :)

Have you tried the Pilot Sport Cup? It's pretty damn good if you start with a lower pressure. On our 996, we start with 28psi front/rear and the car is awesome. It maintains is performance level for a long time, and you can get multiple runs on them. I think at VIR Grande, they ran non stop for 3-4 hours and still had a ton of life left. One of my ex-coworkers (our mutual friend) mentioned that he would rather have the PSC at 235 instead of one of the others at 245. Of course, I've personally not had the opportunity to compare to the others you mention.

Did get to ride (not drive) on some BFG R1's on the Mcoupe. Those things have great grip, but not much warning. They are silent until they let go. Of course these were 4-5 years old so I'm sure they changed just a little. :D

They don't tell me anything about future plans. ;)

The rig was very helpful in determining how the car responded to those specific vertical inputs. It does not "spit out set-ups", and as such is no substitute for sound knowledge of vehicle dynamics. Like any tool, it has to be used properly. It just provides other pieces to the puzzle.

The folks at Penske Technology Center were great to work with as well!

- Dick

I'd like to hear about some of the practical uses for the data gained from the shaker rig, without you guys giving away any secrets, if that's possible.

As far as I'm aware, the Toyo's World Challenge runs are no different from the Toyo's you can buy from Vilven or other distributors.

True story. And Vilven is the ONLY one you're allowed to buy tires from for use in SWC racing. I think it's legal to buy full tread rains from any Toyo dealer, though.

Very interesting. But, has the data resulted in better lap times? Does the car allow drivers to feel more confident? Or more generally, do the data allow for improved car setups?

I don't suppose I can get them to do my E30 before this Sunday's autocross event . . . .

cheers,

Joe Stephenson

You generally figure out some things about the chassis, some things about the way the tires flex, and some things about how springs and shocks work together in a specific package. Helpful? We are generally faster this year than the track records set last year before the 100# weight penalty and weight bias change which we have to battle now - they have reduced it to 50# and 50% recently. If we weren't dealing with a new 6 and TSX, the results would be very telling as far as what it did for us. Instead, we have to look at lap times and remember the good old days when the BMW had power.

Whose shaker rig is it (manufacturer)? One of my best friends writes the software for Roehrig Engineering.

George, I believe the ARC that is now the Penske tech center had setup by Reynard Motorsports, I think they did another one in Indy, as well as that one in NC. There's a lot of nice research places in NC.

Does Roehrig make shaker rigs, I thought they only made dyno stuff?

George, I believe the ARC that is now the Penske tech center had setup by Reynard Motorsports, I think they did another one in Indy, as well as that one in NC. There's a lot of nice research places in NC.

Does Roehrig make shaker rigs, I thought they only made dyno stuff?

To tell you the truth, I don't know if they make them or just sell or use them. I know Doug does programming for them.

I'll have to call him and find out.

Ah, ok.

So whats he program? Who's he work for?

cool

Instead, we have to look at lap times and remember the good old days when the BMW was lighter...

There ya go James, fixed it for you ;)...

BTW, thread title was hilarious...I literally laughed out loud once I realized the joke :).

Ah, ok.

So whats he program? Who's he work for?

I'm pretty sure the program is proprietary.

He works for Roehrig. He has clients in every form of professional racing, including the Ferrari F1 team (Ross Brawn has even called him).

I meant whats the program for, is it running the hardware, or data acq for the shaker, or what? Does he solely work for Roehrig?

Ross Brawn calls everyone :)

I meant whats the program for, is it running the hardware, or data acq for the shaker, or what? Does he solely work for Roehrig?

Ross Brawn calls everyone :)

Yeah, he called me last week, but I was reading the comics so I told my secretary to tell the SOB to piss off. ;)

Running the hardware and the data acquisition. He wrote the data acq software and user interface for Competition Data Systems (CDS) and from that he moved to NC to work for Kurt Roehrig.

He's actually moving to Wyoming soon (his wife is an ER doc and she's already there), but will still work for Roehrig for a while. He and I have been having some discussions about him writing custom interfaces and analysis software for some commonly available amateur and low lever pro data acq. He isn't sure of the market (he's obviously used to the high end pro market), but I've been convincing him there is a real market out there.

The ARC in Indy will cost you about $5k per day for the 7 post, $10k per day for the WT... not sure about the one in NC... plus costs on top of that if you want ARC to give you engineering advice, etc. There are also great seven posts at Ohlins (also in Mooresville) and Dynamic (Toronto, CA).

When you go to the one in Indy, you put your car on the pans and get it all set up, then you can run frequency sweeps, roll sweeps, whatever you want... and it spits an excel sheet back at you with pitch, heave, ride frequencies, and coefficients that are normalized... basically it spits one number back at you and you try to reduce that number through damping/tire pressure/chassis/suspension geometry changes. Essentially, you pay them money and they'll help you find mechanical grip.

I would love to see results of a KNC rig test on a Motorsport chassis... a standard E36, and a standard E36 with a decent cage in it....

Initially, Bruce McCaw (PacWest team owner, bascially he and his brothers started what is now Nextel) and Adrian Reynard commissioned the construction of the ARC in Indy (first rolling road wind tunnel in the USA for racing)... and Bruce still owns some of it, not sure about Reynard.

ARC designed everything themselves, I believe, and wrote all of their own software. Lots of Matlab and VBA stuff.

I would love to see results of a KNC rig test on a Motorsport chassis... a standard E36, and a standard E36 with a decent cage in it....

I will volunteer all the various cars to test and transportation to the NC facility, now who wants to pitch in?

No KNC rig at either of the ARC facilities... the only one I know of is at Pratt and Miller in Detroit... and Dynamic (i would assume) has one

KNC = kinematics and compliance... basically assigns a number to chassis stiffness

If you have the money, I can put you in touch with a friend that runs a KNC rig. There are two there, one big enough for Semi's. Custom built to boot ;)

It doesn't do much shaking, but sure does flex the crap out of a car. :eek:

oh, and if you have to ask how much, you probably don't have the money. :devillook

There are also many places like Arvin Meritor which have some very high end propietary equipment. Some amazing stuff, but if you do not have formal training and experience with that particular "rig" its not worth much money. Its what you do with the numbers thats important, these are all just tools, which mean nothing without a good operator.

I will volunteer all the various cars to test and transportation to the NC facility, now who wants to pitch in?

Would that someone in NC had a K&C rig...

For those interested, I would refer you to SAE paper 2004-01-3547, Using K&C Measurements for Practical Suspension Tuning and Development by Phillip Morse. It was presented at the 2004 Motorsports Engineering Conference & Exposition, November 2004, Dearborn, MI, USA, Session: Vehicle Dynamics - Dealing with the Driver - Part 3.

You can order it here: http://www.sae.org/servlets/productDetail?PROD_TYP=PAPER&PROD_CD=2004-01-3547

I was one of the folks who reviewed the papers prior to the conference.

Session: Vehicle Dynamics - Dealing with the Driver - Part 3.

:lol There are 3 parts on dealing with the driver? Are we that hard to deal with?

Where's the paper on dealing with your engineer? Without engineers, who else would we blame? :stickoutt

:lol There are 3 parts on dealing with the driver? Are we that hard to deal with?

Where's the paper on dealing with your engineer? Without engineers, who else would we blame? :stickoutt

When you win there's noone else to blame but the engineer ;)

BMW puts those numbers out. I had them for the various E36 and E46 models at one point.

:lol There are 3 parts on dealing with the driver? Are we that hard to deal with?

Where's the paper on dealing with your engineer? Without engineers, who else would we blame? :stickoutt


On a serious note:

It actually takes the whole package nowadays. The days of "Fred Hero Driver" overcoming an ill-handling car with shear driving skill are over - save for very low levels of the sport.

It takes an engineer to develop the car (In some mature series where the cars have stabilized and there isn't much development an experienced crew chief might suffice.), a team manager to organize the logistics and (in some cases) set strategy, a chief mechanic/crew chief to organize the car work, a bunch of mechanics to execute it all and prepare the cars, and drivers who arrive motivated and ready to do their best. If any one of those pieces is missing, the team will not do well - and certainly won't win. (There are hybrids, as some drivers are good at setting up their own chassis. On single-car teams, the crew chief or engineer may also serve as the team manager.)

The engineer-driver communication is EXTREMELY important to the whole package. While modern data acquisition systems measure certain parameters, the driver is still the single most important feedback source.

- Dick

BMW puts those numbers out. I had them for the various E36 and E46 models at one point.

Problem with those numbers is how to use them. Every manufacturer puts out their chassis stiffness numbers but if the test set ups are not the same then comparison is moot. Typically, comparison is indeed moot. This became a rather heated discussion during a meeting I was in during the development of the current Mustang a couple years ago. Nothing like silencing a room of Ford Engineers when you quote BMW numbers. Course I knew they would have no idea about comparable test methods. ; )

James, can that machine accept road load data? Seems that would be way more useful that simple sine sweeps. Certainly there is benefit to a sine sweep for for overall development RLD would be the way to go. Bridgestone/Firestone in Akron has some seriously kick ass test equipment for working with complete suspension development..inlcuding tire stiffness. The machine they have is one of a handful in existance, very cool stuff.

So what range of sweep was in the video there? I tried to analyze it with our software from the video but the background noise was pretty high and instrusive, hard to get good data. Sounded liked maybe ~3Hz up to maybe 30 or 40Hz?

Somone else mentioned amplitude lowering as frequency increased. Yes, amplitude is inversely proportional to frequency. Of course I'm speaking in terms of the output of the machine, the response is dependant on the test subject.

There are also many places like Arvin Meritor which have some very high end propietary equipment. Some amazing stuff, but if you do not have formal training and experience with that particular "rig" its not worth much money. Its what you do with the numbers thats important, these are all just tools, which mean nothing without a good operator.

Exactly. These test systems are fairly commonplace around Detroit here, even for race work. Of course the majority of the work is for general road vehicle development. A couple places I worked years ago, including Roush, always threatened to buy one but never made the investment. Well Roush may have by now, the other place did not. Interestingly enough, GM has some of the most fabulous and expensive test rigs I've ever seen. I haven't spent enough time there yet to understand why they seem to be "not taken advantage of" though. I'm keeping my eyes and ears open though.

Whose shaker rig is it (manufacturer)? One of my best friends writes the software for Roehrig Engineering.

There are several different makers of these type rigs. MTS is one of the big ones, fairly common here in Detroit. It's been so long I don't recall the others I've seen. I always remember MTS because they have a test rig for everything. They did some custom (albeit low end) test rig development for us a few years ago for power steering systems. And I had friends who worked there.

Jack, where are you currently working in Mich (e.g. which company)?

What equipment ave you used?

Thanks,
Steve

Problem with those numbers is how to use them. Every manufacturer puts out their chassis stiffness numbers but if the test set ups are not the same then comparison is moot. Typically, comparison is indeed moot. This became a rather heated discussion during a meeting I was in during the development of the current Mustang a couple years ago. Nothing like silencing a room of Ford Engineers when you quote BMW numbers. Course I knew they would have no idea about comparable test methods. ; )


Yes, it would be difficult to compare numbers without knowing the test methodology.


James, can that machine accept road load data? Seems that would be way more useful that simple sine sweeps. Certainly there is benefit to a sine sweep for for overall development RLD would be the way to go. Bridgestone/Firestone in Akron has some seriously kick ass test equipment for working with complete suspension development..inlcuding tire stiffness. The machine they have is one of a handful in existance, very cool stuff.

So what range of sweep was in the video there? I tried to analyze it with our software from the video but the background noise was pretty high and instrusive, hard to get good data. Sounded liked maybe ~3Hz up to maybe 30 or 40Hz?

Somone else mentioned amplitude lowering as frequency increased. Yes, amplitude is inversely proportional to frequency. Of course I'm speaking in terms of the output of the machine, the response is dependant on the test subject.


The rig can accept road load data, but BW didn't have anything useable from 2004. There was no real test/development budget, so the rig test was a major step for them. I had exactly one day, and with that all I could do was get some basic response data. That's what an independent WC team without a major sponsor is up against. Likewise, one can hurt the car when using track data unless one has sufficient experience with the car and rig under those conditions. That would be a future step - given the proper budget.

The test is a sinusoidal sweep with a constant peak speed of 6.3 in/s (160 mm/s), between 0.5 and 16 Hz. As such the input amplitude is reduced as the frequency increases. (Nothing secret about that, but the car's configuration for that run is.) The Penske Technology Center could run any number of other tests, but since one crawls-walks-runs, this was a first step. (Ohlins uses a slightly different test sequence.)

- Dick Golembiewski

Just wanted to point out, or comment on the fact that just because another team, company and/or manufacturer does testing with a differnt method does not mean its wrong, or incorrect. The whole purpose of this type of R&D is to not only get the car to its full potential (and potentially design the components to go beyond its present potential) but its also used to develop the hardware/technique/processes.

The person who wrote the SAE paper quoted earlier has now installed a K&C rig in Salsbury, NC. It is an independent lab, and open to all. The rates aren't too bad, either, with discounts for students, FSAE teams, and first-time users. Website:

http://www.morsemeasurements.com

- Dick

Thanks Dick, that's good info to know for those of us who are local.