ABC of dyno testing for dummies

most interesting to me will be to see how they dyno on the same day, same dyno, same time, same temp etc. Another words, how much more WHEEL HP does a 550 have, and how will the torque differencial add up to better acceleration (hard to believe a 550 wont pull on a 360 after maybe 60 MPH or something)

 

If the real reason for the dyno run is no more than comparison, then just get the same operator on the same dyno doing the same thing as soon as possible after the first time. Most dynos can overlay graphs and print them. Should be a very interesting exercise.

 

If all else is equal (engine temp is the biggie), back to back dyno runs on the same car should be nearly indistinguishable from each other (i.e. when plotted on the same graph they will more or less lie right on top of each other).

 

From my experience in sorting out small improvements in jet engine performance, back to back TO BACK tests are required. Its a pain but you need to see you if get back to the same baseline before you draw conclusions.

Corrections for temperature, maintaining same dyno, operator etc are all required as stated above.

Repeated tests of the same helicopter shaft engine on very very expensive and weekly calabrated dynos with precise procedures and corrections for gov acceptance typically didn't repeat within 0.5%. I would expect more, perhaps a lot more on a typical chassis dyno setup at the local shop. Ask when was the last time it was calabrated and how.

 

I forgot, if it was calabrated recently and found to be off, was the error corrected or accepted as being within "the margin of error".

 

Repeatability is key and there are many factors which can affect this. With the exception of the OEM's and some of the big racers, I really haven't seen many dyno setups that I would trust 100% which is why I'm building a dyno room for my home garage. The room will be used to demonstrate the dyno (I'm a Dynapack distributor) and to independently evaluate many of the aftermarket products being sold to enthusiasts.

The Dynapack chassis dyno (see http://www.ultimategarage.com/dynatech.html for a rather detailed overview) is a hydraulic load dyno that couples directly to the drive hubs. This eliminates 2 important variables right away......tie down strap tension/tire contact (the car need not be tied down) and roller intertia (there is no drum roller...no inertia to). Its also the quietest dyno on the market (no screaming tires), the safest (try standing in front of your 550 while its doing 180mph on the rollers....if the straps broke loose, the car would be launched through the shop wall), and the gentlest on the car (again, no tire wear or tie down straps).

Having the right equipment is one thing but you also need a controlled environment. Most shops simply have big fans blowing heat soaked air around the room. With the right HVAC system, one can properly control temperature, humidity and even pressure (to some extent..through monitoring the room's pressure through the central weather station and using controlled make-up air) in the room.....see http://www.ultimategarage.com/UltGar2g.htm and the 2 pages that follow for my somewhat simplistic approach....I don't have the million dollar budgets that the OEM's do.

By maintaining the same room environmental paramaters before and after the modifications to your car, you should get a pretty good idea of how much impact the mod had on performance. Don't get too hung up on the absolute numbers as they will vary from machine to machine....a Dynapack will read differently from a Dynojet or a Mustang dyno. Also, I would insist on AFR (and "Boost" for you forced induction guys) readings plotted with HP & torque to give a good overview of your performance and setup.

 

I've always been a fan of the simple inertia dynos. Inertia are rpm is all they use which are very easy to measure and will never change, so re-cal is not required, at least for raw numbers. When you use corrected numbers, the air temp, pressure, and humidity numbers need to be correct, so they need calibration to allow comparisons between dynos. But if all the runs are on the same dyno within a few weeks, normally what little drift there is won’t mean anything.

Check you tire pressures if it is a rear wheel dyno. You will see the highest and most consistent numbers if you inflate the tires to the maximum aloud pressure (it’s marker on the tire side wall). Lights, radio, traction control, and AC off.

Make 2 or 3 pulls every time. They should be almost identical, if not you need to fix it before continuing. The 2 CS dyno runs look like 2 different cars, you don’t want to see that. It look like an octane problem, I don’t know if 355s or 550s suffer from it as well, but you may want to use race fuel just to be sure or at least have some on hand if it is available.

Have fun…and post the graphs

 

I will vouch for the effect of tires on a Dynojet dyno run!

I was having problems spinning the tires on my dyno runs and on the street even in 4th gear. So, I sprung for a set of the new BFG gForce TA drag radial tires. I took the car for a drive on the street and these tires are so sticky, they would hold in 3rd gear while the old ones wouldnt even hold in 4th.

Well, I made a return trip to the dyno with nothing changed now feeling that I had the dyno drum slippage problem licked. On the very first run, (on my lowest boost setting), I noted that reported power was down over 10% across the board (i.e. at any rpm there was more or less a 10% difference between runs)! I made another run, and the same result ensued. After thinking about it for a few minutes, it hit me....rolling resistance!

Well, A buddy of mine is a tire engineer for Michelin. (Michelin owns BFG Tires). I called him and asked about this phenomenon. He is friends with the engineer who did the compound design for this tire and that engineer laughed! He said hes surprised I didnt see more of a loss as the rolling resistance of the gForce drag radials is at least 20% more than a "normal" compound tire.

I have a feeling that the Dynojet software must have some "assumed" tire loss coefficient built into the system based on a "normal" rolling resistance. (As it also must assume drum bearing losses, etc). Thus, maybe when a tire is way out of range as far as rolling resistance goes, it ends up showing a power loss.

So, lesson learned there is that one should dyno with the hardest tires possible

I agree with that, especially on an N/A car. If the car is running properly, it has been my experience that back to back to back runs are all very very close in both magnitude and curve shape.

 

Of course, a dyno like the Dynapack wont have to worry about tire losses

 

Check that they can handle your mph/load requirements
i.) Look for a company that has done it before with your type of car
ii.) Know the gear you want to operate in – most close to 1:1 gearing
iii.) Know the correct weight of your car – estimate or measure - one way to get an accurate measurement is through comer balancing scales at a race car shop – 10 min operation

Watch that the car is tied down in a proper way
i.) Don’t let them go at strange suspension components – there is not a lot of load on the straps but you can never be sure – protect whatever frame part the straps are tight to; it will rub
ii.) A good operator will do a quick run and re-check the tightness of the straps
iii.) Check everything yourself!

Think about the octane rating of your fuel
i.) A dyno run is done in a 1:1 gear staring from 2000rpm with the accelerator floored up to max RPM – this is sort of the maximum load you can get – do you have a “good tank of gas”
ii.) Especially during the winter some the quality of the additives is lowered by some gas companies – it might be good to go for a slightly better octane rating

RPM pickup
i.) Make sure that they know how to do it – Ignition, optical, electric fuse… all can be pick-up points for the probe
ii.) A good operator will check with constant RPM

Get all the parameters from the operator
i.) Standard used SAE J1349
ii.) Temperature / humidity correction
iii.) Other variables
iv.) This is also the way to check if the operator knows what he is doing.

Checks you should do:

i.) Race / Sport mode (if you have the F1 gearbox)- Dyno run is done in one gear under full load - automatic switching of gears would mess the result up
ii.) ASR off (or any other stability program) - Slip or stability controls will go crazy if one pair of wheels is spinning and the other not
iii.) A/C off - Sucks a lot of power
iv.) Lights and other devices off - May suck some power
v.) Full tank of gas (makes runs repeatable)
vi.) Tire pressure correct (makes runs repeatable)

You might want to open engine cover to dissipate heat

If you want to be anal (and a good shop will do this anyhow):
i.) covers for drive seat to protect it if operator drives it
ii.) covers for car body areas that are touched by cables, etc.

Procedure
i.) Warm car up to operating temperature (either on the road or on dyno)
ii.) Coast down run
iii.) Max. three runs in a row and take the best of the three
iv.) Cool down cars for an extended time to get down to operating temperatures - this might not reflect itself in the engine temperature but is more to get other parts like cats etc. cooled down – my opinion
v.) Repeate iii.) if you must but results will suffer

You should check after each run temperature of water, oil,.. Check also if no warning lights have come on etc. - Check yourself!

Results that should be printed out in one diagram:
i.) Coast down run over RPM - Tells you parasitic loss of trans, gearbox, engine – claimed to be to low because engine does not drive, i.e. fudge factor needs to be added
ii.) HP rear wheel over RPM - HP = (Torque * RPM) / 5252, i.e. HP is computed from torque - Note that HP & Torque curve have to cross at 5252 RPM or something is wrong!
iv.) Net HP = HP rear wheel + Coast down over RPM - This gives you and indication of what your Flywheel horse power is
v.) SAE corrected HP RPM - This is the HP number corrected according to the selected standard (i.e. SAE) – between different dyno runs this number should stay the same
vi.) Torque over RPM - This is the base number that can be most trusted
vii.) Max Torque, Max. mph and Max. RPM

Get the results in digital format if possible. Resolution of most of the printouts is a shame. Nice to have are measurements over time if you want to analyze the data.

AFR measurement
i.) AFR gives you the best indication of how the engine performs
ii.) AFR measurements in the exhaust tips are not very accurate
iii.) Most cars have a connector for a probe at the end of the header, i.e. pre cat – this is the place to go for
iv.) AFR measurements are only useful if they are logged and you can overlay them over the Torque diagram.

Things to read – both are critical writings and one is defensive – you can find the hype on each of the dyno company Web sites:

Dirt on Dynos by Wiley Davis
http://www.thejauntlist.org/wiley/pub/dyno.pdf

Mazda Statement Regarding Dyno Tests of RX-8
http://www.gtplanet.net/forum/t-30739.html

My general advice is to talk with the dyno operator about all this before the actuall run!

They should have ear plugs - use them

 

"Of course, a dyno like the Dynapack wont have to worry about tire losses "

I guess numbers from a dynopack can't be compared to anything except themselves...it isn't crank hp nor is it really wheel hp because it's not measured though the tires. I guess you're not really supposed to be coomparing anyway though...

 

True enough. Best bet is to just use the same type of dyno each time. The exact same dyno each time is even better.

The previous owner of a car I had a few years ago dynoed it on a Dynapack. I then dynoed it on a Dynojet. The curve shapes werent even the same. Peak power as shown by boh dynos was reasonably close, but one dyno said peak torque was 5000 rpm and the other 3500 rpm! Seat of the pants says the 3500 rpm is closer to the truth.

 

"Peak power as shown by boh dynos was reasonably close, but one dyno said peak torque was 5000 rpm and the other 3500 rpm!"

Sounds like tire slip, I'm guessing the 3500 number was on the dynopack??




thomas_b
Great list! It looks like you covered everything that can be done. I like the check list idea....I always manage to forget something.

 

same with me - another rule along these lines I always find I violated after the fact - Never change two or more things at the same time between runs!

 

With the Dynapack and presumably other models, you can input your car's weight and plot acceleration (in g's or tractive effort). The plot here was a well tuned Porsche turbo weighing 2500 lbs and making close to 600hp at the (rear) wheels. Plot is a 5th gear (1:1) pull....this car accelerates abruptly between 3800-5000 rpms as boost peaks at around 17.6psi @ 4700rpm's. Driver spun the car into the wall at Pocono during its first outing (in the rain)....the graph should have been a warning of how abruptly the boost would affect rear wheel torque/hp.
Image Unavailable, Please Login

 

If I had to make a guess, thats what I would have said too.

Actually, it was the other way around.

Rather perplexing to me, but I suspect that maybe the load rate of the Dynapack was such that the ECU was having a harder time modulating the wastegate solenoid (i.e. overshoot, undershoot, overshoot, etc).

 

Another thing that was slightly perplexing to me (until I thought about it), was that when tire slippage starts on a Dynojet it will actually INCREASE the reading. This fooled me for a bit.

What ends up happening is the "rpm/mph" ratio the dyno is looking at gets skewed when there is slippage in the driveline. (This goes for clutch slippage as well. It also explains why a car with an auto trans will show a huge torque bump right at/around the stall point of the torque converter.) The Dynojet needs a fixed "rpm/mph" ratio during the run in order to produce good data as this ratio is what is used to determine the "effective gear ratio". When that ratio isnt fixed, as when the tires slip, it ends up getting read as extra torque multiplication.

 

Heres what tire slippage looks like on a Dynojet.

 

had to think a little about what you said - the dyno measures the acceleration of a known mass - the drum - by sampling the rotational velocity at the drum shaft and compute the acceleration from the deltas - I believe that is what you mean by "mph" - that should be a very stable and accurate parameter, i.e. torque over "mph"

the dyno uses the RPM that is picked up from the engine to compute torque over RPM - at this point your drive line ratio (effective gear ratio) comes into play - if the tires slip on the drum the RPM to "mph" ratio gets out of wack and you see funny results - I get it

in thinking about it there should be actually two types of slips - tire faster than drum and drum faster than tire (e.g misfire or similar decellaration of engine)

I always felt that when I was scrubbing the results generated by my runs that I was missing data to check plausibility - having also the drum mph besides RPM at minimum would allow to check if not fix things - if the drive line ratio gets out of wack you know that something is wrong

add this to the list:

- get the dyno drum "mph" value for each RPM sample point (and check that it conforms to a fixed ratio)

 

The Dynapack actually measures hub speed and can tell you within 1 rotation accuracy the speed at each drive wheel. This is helpful in troubleshooting final drive problems or even a hanging brake caliper. By knowing hub speed, final drive ratio and gear ratio, the Dynapack calculates engine RPM's (far more accurate than the tach in the car) and by inputing tire diameter, it will calculate road speed (with enough accuracy to calibrate the speedometer).

The chassis dyno (most of them) are good tools in the hands of the right person but they all have their limitations.

 

Have the dyno operator make a single run all the way from zero in first through Redline in 6th. By looking at the shape of the torque curves you can determine the reletive efficiency of each gear. The Tq curve in 1st gear will be considerably reduced that the tq curve in 6th, 1st gear will be lower!

You want the A/F mixture on this 1st->6th run. Since this run gives the engine plenty of time to get its act together, the A/F in 4->6th will show you the control the ECU has over mixture in this section of the power envelope. The A/F reading on the single gear charts will show you how fast the ECUs go from closed loop to open loop and what the mixture does during this transition.

Make sure the engine oil temperature is up to (at least) normal 165-180 temps and stays below 265 dF. (and don't forget to check the oil level!)

Make sure the tires are inflated correctly and fairly new (>1/2 life remains)

Put some good gas in the car, and drive about 50 miles for the ECUs to learn this gas.

You want a dyno graph of you 1:1 gear with all (3) runs on the same graph. The consistency tells you the basic accuracy of the dyno and teh basic repeatability of your car. You want both HP and TQ on this graph A/F if you can get it. For your second graph, you want the 1st->6th run and the A/F. I find TQ more interesting, but have them print TQ on one graph and HP on the other.

 

Did you try here :

http://www.mecacomponents.com

It's where I go when I need to dyno my car !