Three simple questions - pls point me

[maguromic]

Close only counts in  horse shoes and hand grenades. I haven't used the program, but like mentioned in the previous post I think there are better ways to do it.  Tony

[jl222]

  Yea... Bville pro is way off on acceleration data because it assumes one can apply full throttle. its probably
pretty close [doesn't count?] on lower hp cars or bikes.

  I can see how bville pro can be off on acceleration because where is he going to get info about Bville.

  Its hard just to get a room.

  In Ob.. were lucky as far as CD because the stock CD numbers for Camaro is published [.35] and others are also
  I just estimate it at .3 because of lowering and no mirrors. There are some on here that could give a lot more info on this and other things downforce but don't.

  If we had top speed traps at each mile and shut down before the 4th mile, we could figure the coast down

                   JL222


    

  

[lsrengineer]

Lsrengineer:

Your curves came out very nicely.  Looks like Suzuki did a pretty good job with their ratio splits.

From your earlier posts, it would seem that you are embarking on making up a performance prediction worksheet.   I have done this previously for a drag racing application, even involving a torque converter in order to explore ratios, shift rpm and other parameters.  With a little tuning it would duplicate the Racepak data collected on the car and give good ET’s and speeds, as well as trends due to changes.  (Traction limitations were not an issue in this case, but that can be fairly easily handled).

I only mention this because the analysis was based on short time increments, while you had mentioned using distance increments.  I hadn’t tried to use distance increments, but know using time produced a pretty straight forward (non-iterative) set of calculations.  You might want to consider it before you get too deep into it.  Naturally, speed and distance are then easily derived from the acceleration/time results.  Also, if run data is taken on the vehicle, it will most likely be as a function of time as well, making comparison easier.

Including any corrections for lemurs is up to you.

I have the time increment version running, tried various increments and currently using 0.1 seconds, smaller doesn't seem to move things much.  I have to go through the sheet and move the shift points around, looks like the shift points are not that critical but a wide ratio transmission may be more sensitive.  I'll post if the owner doesn't have an issue.

[lsrengineer]

Lsrengineer:

Your curves came out very nicely.  Looks like Suzuki did a pretty good job with their ratio splits.

From your earlier posts, it would seem that you are embarking on making up a performance prediction worksheet.   I have done this previously for a drag racing application, even involving a torque converter in order to explore ratios, shift rpm and other parameters.  With a little tuning it would duplicate the Racepak data collected on the car and give good ET’s and speeds, as well as trends due to changes.  (Traction limitations were not an issue in this case, but that can be fairly easily handled).

I only mention this because the analysis was based on short time increments, while you had mentioned using distance increments.  I hadn’t tried to use distance increments, but know using time produced a pretty straight forward (non-iterative) set of calculations.  You might want to consider it before you get too deep into it.  Naturally, speed and distance are then easily derived from the acceleration/time results.  Also, if run data is taken on the vehicle, it will most likely be as a function of time as well, making comparison easier.

Including any corrections for lemurs is up to you.

Interested Observer,

Here are first pass results of my simulator using 0.1sec time increments.
I need actual recorded engine rpm and wheel speeds to refine too much more.
There are probably some trends that may be interesting to the owners.
The list of assumptions is pretty long so this may have limited value at this point.
For example: I assumed no slip and no tire growth.  There is some of each.
There is some interesting papers I found on traction in gravel and slip ratio, could be something there to use.

[Interested Observer]

Again, good looking results very nicely presented. 
Having fitted the 4/5 mile data from the timing slip, it appears the simulation is a bit optimistic through the mid-range, and results in a pretty long half mile “push off”.  This leads me to a couple of questions. 

Is there any accounting for the surface coefficient of friction?  If not, this might reduce some of the optimism in the early stages and approximate soft/slippery spots and lifting for wheelspin.

Can you “short-shift” in the simulation, as you most likely would in reality?  That is, can the shift rpm’s be specified for the lower gears and then just let the simulation run?

(I would be dubious of the value of rolling resistance predicted for gravel.  I think you will find that performance on the salt is governed more by the coefficient and wheel slip/growth than displacement of the roadbed).

[lsrengineer]

Again, good looking results very nicely presented. 
Having fitted the 4/5 mile data from the timing slip, it appears the simulation is a bit optimistic through the mid-range, and results in a pretty long half mile “push off”.  This leads me to a couple of questions. 

Is there any accounting for the surface coefficient of friction?  If not, this might reduce some of the optimism in the early stages and approximate soft/slippery spots and lifting for wheelspin.

Can you “short-shift” in the simulation, as you most likely would in reality?  That is, can the shift rpm’s be specified for the lower gears and then just let the simulation run?

(I would be dubious of the value of rolling resistance predicted for gravel.  I think you will find that performance on the salt is governed more by the coefficient and wheel slip/growth than displacement of the roadbed).

All good comments and questions:

1 - I played with the push off distance and speed at push off - I need some actual data to figure out what is going on there.  The simulator is optimistic by a considerable amount at 1/4 and mile 1.  A lot of different things could be going on, e.g. our engine is not matching the power curve from others dyno results, the driver is taking some time to get settled in the early parts of the course.  I don't know.
2 - I can short shift the simulation and see what happens, when I was playing with the shift points it didn't look to change things much.
I shift when the force in the next gear is larger than the force of gear I'm in.  For this hp curve it is well past the 13.2k rpm hp peak.  I'll try short shifting by 500 rpm at each gear and see what happens.
3 - coefficient of friction - I'm still thinking how to handle it.  I've been reading some good papers on how the static and dynamic COF changes with speed/etc. for different surfaces.  Gravel was interesting not for rolling resistance but for COF at various slip rates.  I think salt might be somewhat similar to a well packed gravel road.  I still need to find a team willing to show me their driven and undriven log data.
4 - for our car we don't think we are getting a lot of slip, low HP (~125), good weight on the rear tire,  the tire shows very little in terms of looking like things are sliding and we don't hear it and the driver hasn't mentioned he feels anything like that.
5 - tire growth, I need to re-read the posts and see if I can assume the amount at the speed we are running, I assume the growth is equal to speed ^ 2.  I mentioned this to T. Burkland and he didn't have data on our tire.
6 - rolling resistance - good question, right now assuming the rolling resistance is in the dyno values.  There are some good papers on this as well, the rr might go down when the tire grows if contact patch gets smaller, rolling resistance is somewhat a function of the contact patch size and location.

Thanks for the comments.

[Glen]

Close only counts in  horse shoes and hand grenades. I haven't used the program, but like mentioned in the previous post I think there are better ways to do it.  Tony

Lots of seat time will help!

[Rex Schimmer]

Isrengineer,
I just went through this thread and in your Sept 30 posting you state "Should I find the constant that makes the 1/2 A Cd rho v squared line cross the 6th gear line at 173 mph?" Just a note aero horse power is a function of the cube of the velocity not the square. Force is a function of the square of the velocity.

Also a note on gearing an finding the correct gear vs. the engine power curve. This year at Speed Week the no. 919 J gas streamliner of Briant-Wright and Sparanza reset the record from 213 to 219 mph using a 600 cc engine, which is giving away 150 ccs to the class limit.  Early on in the meet they were running around 180 and it seemed they might have found their limit. Then they started to shorten the final ratio and let the engine really turn some rpm next thing you know they had the record! I am sure that they also did some additional tuning but they told me that changing the gearing and letting the engine turn faster was the major change. Obviously this small engine needs rpm to make horse power and they needed to find the correct gear to allow the max hp to cross their max speed line and it was a record. As they did take this from one of Jack Costella's cars that was running a 750 cc engine I would say it was one of the really outstanding performances of the meet.

Just a note on the car that did hold the record, apparently Jack missed the part about needing to have two wheels driven,  4.D in the rule book, and was not allowed to run.

Rex

[SPARKY]

Tractive effort )  TE = TQ x TR  (trans ratio) X RA (rear ratio) X TC  tire correction

[hotrod]

Is there any accounting for the surface coefficient of friction?  If not, this might reduce some of the optimism in the early stages and approximate soft/slippery spots and lifting for wheelspin.

Due to gear multiplication of torque you are power limited in acceleration at the lower gears. Any time the tractive effort gets close to the available traction, (weight on drive wheels x coef of friction) wheel spin and forced throttle lift become an issue.

If you calculate the peak engine torque x transmission gear ratio x differential gear ratio and divide by the rolling diameter of the drive wheel in feet, you get peak force applied at the tire contact patch. If that force gets close to the weight on the drive wheels x coeff of friction / number of drive wheels, wheel spin is almost guaranteed.

A few years ago I did a spread sheet of major production cars that were notorious for being prone to wheel spin on throttle application, and found that there was a very consistent relationship as stated above. Some cars like the 409 chevy were well known to be capable of doing a power induced spin out just pulling out of a driveway and trying to accelerate up to traffic speed, or simply stabbing the throttle at traffic speed could light the tires. When I checked the gear multiplication vs engine torque I found all those cars came very close to the maximum thrust that could be applied to the tire patch on good clean pavement. Any small loss of traction due to a bump, dirt on the pavement, an oil or wet spot and you would light the tires.

As many of the drivers know out on the salt you have to really tip toe through the first few gears on the high power cars, especially as they come on boost if turbocharged to avoid an embarrassing starting line spin.

If you don't already have it included in your simulation, you should apply a similar applied power limit in the low gears based on an assumed friction coeff in the lower gears, and weight on the drive wheels since you know it is impossible to apply full engine power in those gears due to wheel spin/traction issues. Real drivers on a real traction limited surface will tend to apply power carefully not to exceed that traction limit and will almost always be conservative if they wish to avoid wheelspin, so a limit like 80% of the theoretical traction limit is probably a good place to start as most drivers will back off as soon as they sense wheel spin to well below the limit, and then walk back into the throttle.

Larry

[Interested Observer]

Due to gear multiplication of torque you are power limited in acceleration at the lower gears. Any time the tractive effort gets close to the available traction, (weight on drive wheels x coef of friction) wheel spin and forced throttle lift become an issue.

If you calculate the peak engine torque x transmission gear ratio x differential gear ratio and divide by the rolling diameter of the drive wheel in feet, you get peak force applied at the tire contact patch. If that force gets close to the weight on the drive wheels x coeff of friction / number of drive wheels, wheel spin is almost guaranteed.

Hotrod -- first paragraph - correct thought but wrong word.  Should be "traction limited".

Also, second paragraph - you may want to re-think the "number of drive wheels" divisor.

Your last paragraph is a good description of the situation.

[lsrengineer]

Is there any accounting for the surface coefficient of friction?  If not, this might reduce some of the optimism in the early stages and approximate soft/slippery spots and lifting for wheelspin.

Due to gear multiplication of torque you are power limited in acceleration at the lower gears. Any time the tractive effort gets close to the available traction, (weight on drive wheels x coef of friction) wheel spin and forced throttle lift become an issue.

If you calculate the peak engine torque x transmission gear ratio x differential gear ratio and divide by the rolling diameter of the drive wheel in feet, you get peak force applied at the tire contact patch. If that force gets close to the weight on the drive wheels x coeff of friction / number of drive wheels, wheel spin is almost guaranteed.

A few years ago I did a spread sheet of major production cars that were notorious for being prone to wheel spin on throttle application, and found that there was a very consistent relationship as stated above. Some cars like the 409 chevy were well known to be capable of doing a power induced spin out just pulling out of a driveway and trying to accelerate up to traffic speed, or simply stabbing the throttle at traffic speed could light the tires. When I checked the gear multiplication vs engine torque I found all those cars came very close to the maximum thrust that could be applied to the tire patch on good clean pavement. Any small loss of traction due to a bump, dirt on the pavement, an oil or wet spot and you would light the tires.

As many of the drivers know out on the salt you have to really tip toe through the first few gears on the high power cars, especially as they come on boost if turbocharged to avoid an embarrassing starting line spin.

If you don't already have it included in your simulation, you should apply a similar applied power limit in the low gears based on an assumed friction coeff in the lower gears, and weight on the drive wheels since you know it is impossible to apply full engine power in those gears due to wheel spin/traction issues. Real drivers on a real traction limited surface will tend to apply power carefully not to exceed that traction limit and will almost always be conservative if they wish to avoid wheelspin, so a limit like 80% of the theoretical traction limit is probably a good place to start as most drivers will back off as soon as they sense wheel spin to well below the limit, and then walk back into the throttle.

Larry

No issues here.  I'm in the process of putting this in.  Yes, if tractive effort exceeds the COF*weight then you will start spinning.  This puts one on the dynamic COF chart.  Got any ideas on the static COF?  Should we start with .5?  I'll pull up the papers on this and re-read to see if there is a better number to start with.  I think I read somewhere here that salt has about 50% of asphalt.  I think asphalt is 1.0.

[Interested Observer]

Hotrod,
While in the shower it dawned on me that your TE calculation used the tire diameter, not radius, and therefore the "number of driven wheels" does need to be included as you have done (assuming two wheels).  I hereby retract and apologize for the earlier divisor comment.

[Interested Observer]

lsrengineer,
COF = 0.5 is a good place to start, maybe 0.6 or 0.7 for good conditions.  Bearing in mind that the COF will vary a considerable amount depending on course conditions.  Sometimes the salt is like clean concrete and sometimes like polished concrete with sand scattered all over it.  Sometimes it’s slightly mushy.  Most of the time it is some sort of an average of these.

I’m not sure worrying about the dynamic COF is worth the effort.  As soon as a dynamic COF comes into play, the driver is going to attempt to eliminate it.  If the analysis does the same the dynamic COF, in effect, will hardly exist.  (I’m assuming the analysis is not going to be tasked to predict how many times the car will pirouette before coming to rest).

[wobblywalrus]

The curves show maximum power application in all gears.  This might be a good assumption for paved surfaces.  It would be unusual to apply maximum horsepower in the lower gears on the salt.