Chassis Dyno Information / Guide / Myths!

[darrenjlobb]

Afternoon all


I've been meaning to write this for a little while, but have not got around to it, so here goes...

As most people will know, I've fairly recently installed and setup a Chassis Dyno into my workshop for tuning purposes etc, and it has to be said, I have learnt a huge amount about truths and tales of "dynos" in the process and felt it would be good to share my view on these matters.

First and foremost, when most people are talking about a "dyno" they are most likely referring to a Chassis Dyno, that is a Dyno that you can simply roll a car onto, strap down, and test. There are variations of this regarding direct hub driven dynos, twin roller and single rollers dynos, but any Dyno that is reading torque / power via the car itself is seen as a chassis dyno.


On contrast  to this, there is the "Engine Dyno", this is a Dyno that is hooked DIRECTLY to an engine, on a test stand, to read raw output from the engine itself, with no other losses connected.

Now as most people are fully aware, the disadvantage of running a Chassis Dyno is down to the fact that you are transmitting the original power from the engine, to the roller shaft, however on the way there it is subjected to several places of "losses"...including things such as the transmission, final drive / diff, drive shafts / wheel bearings and wheels / tyres....For this reason, the power "measured" at the rollers will obviously be lower than at the flywheel of the same engine, as power is being lost in these various areas, by varying amounts between vehicles / testing conditions etc.

However, I'm willing to bet that "most" of the people on here who have had there car ran on a Dyno were given a graph at the end, which likely reported to be " @ the flywheel / crank "  or if you were lucky, both flywheel and wheel horsepower...Obviously worth noting at this point that whp/wheel horsepower is the powered measured directly at the rollers after losses, and flywheel hp is measured directly at the engines crankshaft on an engine Dyno.... So you ask, how are they able to tell me a flywheel figure when they were measuring at the rollers.... Well that is a very good / controversial question...  

Different Dyno manufacturers have different views on how to go about this, some using "fixed" losses dependant on the type of vehicle etc (I.e. Dyno Dynamics units) and some use the so called "coast down losses" system, where by upon lifting off at the max rpm point, you depress the clutch and allow the rollers inertia to drive the wheels / shafts / tranny down to a stop, and by knowing the rollers moment of inertia the software is able to detect the "losses" at each rpm point...or in theory.... and some simply refuse to even try to do the impossible in the first place and read raw WHP figures.

HOWEVER...both of these systems are entirely flawed... Using fixed losses although may by luck give you roughly correct figures for some cars etc, can equally be way off for other cars tested in similer conditions... For example, from my own experience, there are so many factors that can effect the "losses" that are simply totally ignored using this method of flywheel power generation, say you tighten the front straps down a bit tighter on one car than another, or the cars tracking is slightly out, or the tyre pressures are different, or a brake / wheel bearing is dragging / gearbox oil temp is different....ALL of these things effect the losses, and will be entirely ignored by a static loss...see the problem here?   


So then we look at the more common method of trying to calculate this, the "coast down / run down"....Many of you will likely have been to a dyno day where you here the cars screaming to the red line, and then allowed to "free wheel" at its own rate down to a near stop...this is doing a coast down to calculate the losses going on... Great......Right?!.... Well sadly no..... Although this technique is at least attempting to somewhat calculate what is going on, sadly, its doing it in totally un realistic conditions...for example.. Imagine inside your transmission, the geared are meshed, and turning, do you really suspect that the losses are the "same" while its freely coasting to a standstill, compared to when its having 300hp rammed through it? No....didn't think so! Likewise for the tyres on the rollers...Take a look at the tyre deformation when the car is on full boost / accelerating compared to when its coasting down...I think I'm fairly sure which is creating more heat / power losses!  Therefore, this technique although at least making an effort, will still never be accurate... And that isnt even mentioning "intentional" interactions from the operator to "skew" results..such as dragging the brakes slightly on the run down to increase the losses to make power gains look more impressive etc.

I think the point I'm trying to make here, is ANY "flywheel" figures you are given from ANY dyno, take with a pinch of salt, as at the end of the day, a chassis dyno measures power AT THE WHEELS....and NOT at the flywheel....ANY kind of cross calculation is simply a "calculated GUESS" at the best of times...and really dosnt mean alot in reality. 

You will notice most other places (The states / Aus etc etc) never work in flywheel power, everything you read is whp...and for good reason...Sadly the UK is self obsessed with "pub talk" flywheel figures as these are always higher, and mean more....but in reality, its just nonsense, and any professional engine shop will tell you exactly the same.

The reason I felt I should right this is I read so often online people taking there cars to dyno days / tuning etcetc, and coming home happy / disappointed with the "numbers"...and then comparing them to there "mates" run in his car that was tested at some other dyno...And taking it all for gospel that his car is 20hp or whatever less than his mates...when in reality, it could honestly be 10-20hp MORE.....The truth of the matter is you simply cant compare one dyno to another, there are 101 different variables before you have even started.

A chassis dyno is a fantastic tuning tool, and 100% repeatable once setup / warmed up and running, which is exactly what you need for tuning a car... But to then take your car to another dyno a few weeks later and expect it to make exactly the same is sadly not the case / realistic, and I really think its important people realize this, as all the big dyno names fool you into thinking these numbers are entirely correct and couldn't possibly be wrong, but in fact, there really just a calculated guess.

I'm in no way saying that all UK dyno companies are wrong, of course there not, none of them are any more wrong than any other one, but what I am saying is take it for what it is, and bear in mind unless your looking at whp, Its just going to be a close guess, and could vary considerably if you were to actually engine dyno your motor.


And lastly, another thing that is worth noting, as I have seen a few people dissecting dyno graphs disputing the shown "transmissions losses" and then deciding because they have seen someone elses car have higher losses, that theirs must to, and to re calculate there figures based on this new %....This is also entirely wrong. The entire transmission losses thing would also likely surprise you, for example, most people run with this online Myth that FWD cars have around 15% tranny losses, and RWD cars run around 5% more, give or take....
In reality, the gearbox itself looses very little, in fact in most cases less than 5% box / diff combined...Think about it, if your car has 200hp, and its putting 15% of it into the gearbox as losses, thats like pumping 22kw's of heat / noise into the gearbox, the thing would be melting after a short shopping trip! 
In reality, when your car is on a chassis dyno, most of the losses, are between the tyres and the rollers themselves...This area equates to the biggest amount of total "losses" not only up to 15%, but in alot of cases, quite alot more....

So why am I making this point? Well..like I said at the start...chassis dyno's come in various forms...single roller...twin roller...smooth rollers...knurled rollers...strapdown tension / weight of car...tyre pressures / temps..all of these factors have a massive effect on the "tyre to roller " losses...and therefore the realistic driveline losses will vary massively from one dyno to another based on this alone, so you simply cant look at someone elses "losses" figure and translate it into your own graph etc... 


Until having my own Dyno, I used to see these figures in the same way you likely do, and given what you read online, I wouldn't blame anyone for thinking so, but in reality, the truth is very different from my findings!

Bit of a long thread, but hopefully it might help clear up alot of confusion / rubbish you read online, and also give you a better idea on how the dyno is generating the numbers it gives you in the first place.

Of course everything is open to discussion, of which is most welcome as its an interesting topic, so if anyone is still reading and feels like it, feel free to discuss

Thanks for reading anyway

---

Goodness that was a long / boring text post, some thread related french car photos to ease the mind!!!

[cully]

good write up darren

[Niall]

Awesome post Darren!

[bashbarnard]

Good informative info. I knew a few bits but no where near as much as that!! Always good to have people on the forum willing to share knowledge. At the end of the day that's what its about.

[Damo]

Very interesting read this

[Dum-Dum]

Great information. I'd always been told the losses were about 15% for FWD, 20% for RWD and 25% for 4WD but I always seem to end up a little over that when they do a coast down.

I suppose if you want a fair comparison you need to keep going back to the same operator at the same dyno. The UK is never going to go to talking about WHP figures as there is too many willy wavers about.

[toseland]

which sticker adds thje most ~BHP darren? lol

[Rippthrough]

I suppose if you want a fair comparison you need to keep going back to the same operator at the same dyno. The UK is never going to go to talking about WHP figures as there is too many willy wavers about.

The problem is WHP figures are also just as variable as flywheel figures and aren't an accurate report of what actually does get to the road either. As always, a dyno is a tool for tuning, not really for comparisons between setups, even though it gives you a rough idea.

[darrenjlobb]

I suppose if you want a fair comparison you need to keep going back to the same operator at the same dyno. The UK is never going to go to talking about WHP figures as there is too many willy wavers about.

The problem is WHP figures are also just as variable as flywheel figures and aren't an accurate report of what actually does get to the road either. As always, a dyno is a tool for tuning, not really for comparisons between setups, even though it gives you a rough idea.

What Ripp said ^^

The WHP are also affected by things like roller type / tyre pressures / type of tyres and temperatures etcetc, obviously as this is the raw reading after all  said things are removed.

There is no reading you can take from one dyno, to another and directly compare, not with any form of accuracy anyway, Its a very useful tuning tool, which can also give you a rough idea of your power output, but rough at best imho.

[Rippthrough]

Put it on the road and press the pedal, that gives you the best idea

[Dum-Dum]

I was talking about using the same dyno at the same place for repeatability of WHP figures. Most good operators set your tyre pressures before you run

[darrenjlobb]

 Put it on the road and press the pedal, that gives you the best idea

I wouldnt agree entirely...Or at least since having a rolling road, I would certainly never go back to tuning using the road as a "test medium"...For most modifications you havent a clue what is going on / what you have done! But yeah 1/4 mile strip gives you as good as idea of anything with regard to rough power if you know the vehicles weight.

I was talking about using the same dyno at the same place for repeatability of WHP figures. Most good operators set your tyre pressures before you run

Yes this is correct, Using the same dyno for comparisons on the same car is pretty reliable, obviously small changes can occur day to day and with the strap down, on the whole they are comparable. And yes tyre pressures should always be checked.

[Rippthrough]

 Put it on the road and press the pedal, that gives you the best idea

I wouldnt agree entirely...Or at least since having a rolling road, I would certainly never go back to tuning using the road as a "test medium"...For most modifications you havent a clue what is going on / what you have done! But yeah 1/4 mile strip gives you as good as idea of anything with regard to rough power if you know the vehicles weight.

Well, for testing and modifications you can't beat it, dyno makes things far easier and more consistant - but the most reliable way to find out how it goes down the road - short of a chassis rig - is to go down the road

[r3k1355]

So what stuff can you do to cheat the dyno then, if for example you wanted to wind a mate up or something.

You mentioned the tracking?
Whats the best alignment to have on the dyno wheels??

[bashbarnard]

So what stuff can you do to cheat the dyno then, if for example you wanted to wind a mate up or something.



You mentioned the tracking?

Whats the best alignment to have on the dyno wheels??

Bigger turbo. Lol

[Dum-Dum]

Tyres pumped up and straight tracking id guess.

Narrow tyres and light wheels probably help too and possibly some sort of change in gear oil.

[darrenjlobb]

Depends on the type of dyno to what effects it, but little you can do as the owner, its more what the operator can do to skew it...mainly by skewing the loss's coastdown or temp sensors etc for the "correction" ...

[Fooby]

Thanks for providing this really useful information Darren.

[Pebbles167]

Interesting read, smart lad!

[Matt-Rallye]

 Put it on the road and press the pedal, that gives you the best idea

I wouldnt agree entirely...Or at least since having a rolling road, I would certainly never go back to tuning using the road as a "test medium"...For most modifications you havent a clue what is going on / what you have done! But yeah 1/4 mile strip gives you as good as idea of anything with regard to rough power if you know the vehicles weight.

I was talking about using the same dyno at the same place for repeatability of WHP figures. Most good operators set your tyre pressures before you run

Yes this is correct, Using the same dyno for comparisons on the same car is pretty reliable, obviously small changes can occur day to day and with the strap down, on the whole they are comparable. And yes tyre pressures should always be checked.

This is a fantastic write up with alot of useful information but on the highlighted point of 1/4 mile times to indicate horsepower output that just doesnt work in practice.

For example:

If you take a 150bhp 106 gti up the strip with a standard gearbox you can expect a 15.2 at best.

Now take that same 150bhp in the same 106 but with an S1 rallye gearbox (much closer ratio) and you can pull a 14.2 with a decent launch! That 1 second loss might not sound like much but in the world of drag racing it is everything.

I do have some questions about the figures obtained from the RR day i did in hereford as some dont seem to add up but others seem to make perfect sense. just wondering if there can be a reasonable explanation for it?

[zx_volcane]

Great writeup, hopefully give people an insight behind the all important bhp figures

Also loving the ratio of ZXs to other marques on the dyno

[Piggy]

great write up darren.

It kind of summarises a lot of research I have found on dynos and experienced myself over the years but in one decent post! nice work!

[darrenjlobb]

With regard to 1/4 mile times, if your gearbox is suitably matched to your engine (keeps you on good power 99% of the time) you would be surprised how accurate working out power this way is... The yanks very often use this method to cull out the "made up" figures people blurt about...conditions have to be correct etc, but if the car is hooked, and on power, it can be surprisingly accurate..

[Dum-Dum]

I think a combination of dyno, 1/4 mile and 30-70 in gear gives a rounded opinion of how powerful something actually is.

[Redordead89]

   
If you have a TU....