EVODYNAMICS Ultimate Suspension Data Thread
#1
Former Sponsor
Thread Starter
iTrader: (56)
Join Date: Sep 2004
Location: DFW, TX
Posts: 2,701
Likes: 0
Received 0 Likes
on
0 Posts
EVODYNAMICS Ultimate Suspension Data Thread
First, this thread is to make people aware of how the Evo suspension TRULY acts. Nearly every thread I have read concerning shocks and springs is filled with assumptions. Granted there are a few people with the resources to actually measure stuff, but most don't tell what they find.
Second, this took a great deal of our time and money (correct equipment, sensors, DAQ, brackets made, wired and calibrated) to get to the point we are confident in our findings. So PLEASE, give us a like on Facebook at Evodynamics and My Shop Assist. And be sure to tell any shops you know of about My Shop Assist and how much better their shop will be with it
Third, thanks to Josh for letting us use his car and DAQ as the test vehicle!
With that all said, the test vehicle is a 2006 evo 9. Its total weight is 2950 with driver, wet. Here are the corner weights without the driver:
I have the weights for with driver, but no picture (the WITH driver data is what we will use). lol.
Next, the suspension on the car is AST 5200 coilovers with Swift 10k/12k springs. The rears also have a helper spring.
To ensure the springs were as advertised, we visited Vorshlag Motorsports and Terry was generous to squish the springs on his spring tester. Here are the results:
The measured spring rates were slightly less then the labeled rates. We saw 9.7 kg/mm for the fronts and 11.6 kg/mm for the rears.
From there, we used the MOTEC linear potentiometers to measure the shock position. We have 4 of these, one for each shock, with CNC water jet cut brackets to hold them parallel with the strut. Richard drew them up in CAD and had them cut for us.
We are using a Race Technology Dash2 digital dash and their DL1 data logging box. All 4 shock pots can be recorded at 100hz along with GPS, speed, and accelerometer data.
after a few months of being too busy to actually do any testing, we got some great results with just the rear sensors at the first NASA race of the year. With Josh setting the TT2 track record, we know the laps were good data to use.
But to begin this tech article, we want to keep it simple. Our friends at AST-USA here in Plano, TX were able to get our shocks on their Roehrig shock dyno and ran through a few settings. We elected to only dyno one front and one rear because I was paying to do this
This data will be used in our analysis later on.
For our first test, we took the car at 60MPH down a straight, moderately bumpy road for 100 seconds with the shocks set to full stiff. We then repeated the same stretch of road with them set to full soft. The data generated tells us several things.
Here is the shock positions with respect to time, and the speed plot as well (in green):
Full soft:
The plots aren't super useful as is, but you can see the amplitude of stock travel down this road (max travel of about .1in).
Now we are more interested in the stock rates so we can see where on the shock dyno the dampers are operating. We took the derivative of the position with respect to time and got this cool plot:
The cool thing about this plot is that we can see the dampers are barely moving when going in a straight line and holding speed. Max shock rates are .2in/sec. Again, this plot is hard to read, so we created a histogram from the data and plotted full stiff over the full soft:
With this plot, it is obvious the settings on the dampers make a significant difference in the response of the shock (duh). At full soft, the struts spent approximately 20% less time at the 0 rate position, and that time was distributed pretty evenly throughout the range of the measurements (i cut the plots at -.2 and .2 in/sec).
Over the next few days I will add to this post with adjusting just the compression or rebound independently. I will also add my simulation data to demonstrate what damping ratios we are looking at.
Before people start clamoring about how these data points BARELY make it onto the shock dyno graph (.2 in/sec out of a 10in/sec range), we have data from the track and it uses more of the range. Surprisingly, a lot less then people guess on here. Our goal of this post was to show a proof-of-concept on a straight road before we go to the track again and take measurements there.
(more to come)
Second, this took a great deal of our time and money (correct equipment, sensors, DAQ, brackets made, wired and calibrated) to get to the point we are confident in our findings. So PLEASE, give us a like on Facebook at Evodynamics and My Shop Assist. And be sure to tell any shops you know of about My Shop Assist and how much better their shop will be with it
Third, thanks to Josh for letting us use his car and DAQ as the test vehicle!
With that all said, the test vehicle is a 2006 evo 9. Its total weight is 2950 with driver, wet. Here are the corner weights without the driver:
I have the weights for with driver, but no picture (the WITH driver data is what we will use). lol.
Next, the suspension on the car is AST 5200 coilovers with Swift 10k/12k springs. The rears also have a helper spring.
To ensure the springs were as advertised, we visited Vorshlag Motorsports and Terry was generous to squish the springs on his spring tester. Here are the results:
The measured spring rates were slightly less then the labeled rates. We saw 9.7 kg/mm for the fronts and 11.6 kg/mm for the rears.
From there, we used the MOTEC linear potentiometers to measure the shock position. We have 4 of these, one for each shock, with CNC water jet cut brackets to hold them parallel with the strut. Richard drew them up in CAD and had them cut for us.
We are using a Race Technology Dash2 digital dash and their DL1 data logging box. All 4 shock pots can be recorded at 100hz along with GPS, speed, and accelerometer data.
after a few months of being too busy to actually do any testing, we got some great results with just the rear sensors at the first NASA race of the year. With Josh setting the TT2 track record, we know the laps were good data to use.
But to begin this tech article, we want to keep it simple. Our friends at AST-USA here in Plano, TX were able to get our shocks on their Roehrig shock dyno and ran through a few settings. We elected to only dyno one front and one rear because I was paying to do this
This data will be used in our analysis later on.
For our first test, we took the car at 60MPH down a straight, moderately bumpy road for 100 seconds with the shocks set to full stiff. We then repeated the same stretch of road with them set to full soft. The data generated tells us several things.
Here is the shock positions with respect to time, and the speed plot as well (in green):
Full soft:
The plots aren't super useful as is, but you can see the amplitude of stock travel down this road (max travel of about .1in).
Now we are more interested in the stock rates so we can see where on the shock dyno the dampers are operating. We took the derivative of the position with respect to time and got this cool plot:
The cool thing about this plot is that we can see the dampers are barely moving when going in a straight line and holding speed. Max shock rates are .2in/sec. Again, this plot is hard to read, so we created a histogram from the data and plotted full stiff over the full soft:
With this plot, it is obvious the settings on the dampers make a significant difference in the response of the shock (duh). At full soft, the struts spent approximately 20% less time at the 0 rate position, and that time was distributed pretty evenly throughout the range of the measurements (i cut the plots at -.2 and .2 in/sec).
Over the next few days I will add to this post with adjusting just the compression or rebound independently. I will also add my simulation data to demonstrate what damping ratios we are looking at.
Before people start clamoring about how these data points BARELY make it onto the shock dyno graph (.2 in/sec out of a 10in/sec range), we have data from the track and it uses more of the range. Surprisingly, a lot less then people guess on here. Our goal of this post was to show a proof-of-concept on a straight road before we go to the track again and take measurements there.
(more to come)
Last edited by KevinD; Feb 14, 2014 at 01:16 PM.
#7
Evolved Member
Trending Topics
#8
EvoM Guru
iTrader: (4)
before people start clamoring about how these data points BARELY make it onto the shock dyno graph (.2 in/sec out of a 10in/sec range), we have data from the track and it uses more of the range, but surprisingly a lot less then people guess on here. our goal here was to show a proof of concept on a straight road before we go to the track again and take measurements there.
#9
Evolved Member
iTrader: (10)
#10
Former Sponsor
Thread Starter
iTrader: (56)
Join Date: Sep 2004
Location: DFW, TX
Posts: 2,701
Likes: 0
Received 0 Likes
on
0 Posts
(and for the record, my background is in modeling and simulation + controls for the defense industry. i've worked on some of the most advanced ground vehicles on the planet)
1)laser ride height
2)shock position
3)load cells
with the load cell being the worst at measuring small changes in shock compression
I always thought it would be useful to have a few more closely-spaced points in the lower velocity regions of the shock dyno plot and a few less points in the higher velocity regions. It would be quite helpful to see the detail around the knee, but past the knee the response doesn't change nearly as much from point to point.
#12
Evolving Member
iTrader: (4)
Join Date: Jun 2007
Location: Dallas, TX
Posts: 313
Likes: 0
Received 0 Likes
on
0 Posts
This the sensor you are using?
http://www.pennyandgiles.com/Product...or-MLS130.aspx
http://www.pennyandgiles.com/Product...or-MLS130.aspx
(Bottom of page)
http://www.motec.com/sensors/position/#MoTeC%20Linear
#15
Wheel Rates
Was thinking about building a simple simulation but one would need wheel rates. The spring rates are a start but not very useful. I have some ideas to measure your wheel rates with your shop tools. Center of gravity would also be helpful but measuring it would be more difficult where an educated guess would be nearly as good.