Building an engine dyno.

[Trevor]

It is impossible to measure the torque delivered individually by each cylinder. Reciprocating weight is specifically built into the engine, as a means of eliminating the very factor involved.

A very simple principle is being applied and the application in practice is surely self explanatory. Once the engine is braked, i.e. held back from accelerating, at any given RPM, the force delivered by the brake via a lever, exactly constitutes torque. The length of the lever and the force delivered provides an exact means of calibration. All else is secondary.

It should be noted that variations in the performance of the brake from time to time, as a result of temperature or whatever, have no affect on measurements. All that is necessary is to hold a steady state sufficient for each measurement to be recorded. The brake has only to resist the available torque for that given short time, it does not matter how this is achieved or how the accrued power is dissipated.

In respect of engine tuning, comparative measurement is the key requirement and repetitive accuracy over an extended period is absolutely essential. Complication is a disadvantage.

[Dessertrunner]

Thanks guys for the input it is a real help, I think the project is a bit at the limits of my skill so I need all the input everyone can offer. Harvey that sounds like a great way to do the run.

My reason for testing EGT was more to find the differences in power in each cyclinder is producing. I think it will be a case of read them and see if there is a major difference between each one if so go loooking why.

Okay are you guys siting down I am wondering if its possiable to run one cyclinder for the big first stages of development. Have the other 5 turned off and maybe plug out valves removed. The logic is that I don't have to spend 6 times the money, for example 6 different ram pipes for each trial.
So what do you think is it possiable or just a dumb idea?
Tony

[Dessertrunner]

Current design. The PDF maybe black just click on it and the image will show. Remember its 3d so you can spin etc.
Tony

[Trevor]

Originally Posted by Dessertrunner

---------
Okay are you guys siting down I am wondering if its possiable to run one cyclinder for the big first stages of development. Have the other 5 turned off and maybe plug out valves removed. The logic is that I don't have to spend 6 times the money, for example 6 different ram pipes for each trial.
So what do you think is it possiable or just a dumb idea?
Tony

Kia ora Tony,

Not a dumb idea, particularly in view of the paper you introduced in the throttle body thread. However as pointed out there, at the engine speeds and valve timing you are working with, not altogether practical.

You are involved with a situation where several induction and exhaust tracts are of necessity, arranged to take advantage of gas movement occurring in adjacent cylinders, and therefore all six must be treated as part of a single system. A single detached cylinder would involve quite different parameters. You are stuck with the cost of a six pack.

Trevor.

[Dessertrunner]

Hi Trevor,
I think it would give resonable data as the intake will be ITB and the exhaust will be extrator but should have some simlar bases.
Tony

[Boxersix]

Tony, I took a handfull of pictures of our dyno for you. I've just got to DL them from the camera and UL them to photobucket or picasa later tonight. Wife and I are going out for a bit right now.

I'll post them here later on for all to see since they are on-topic afterall

[dynomatt]

CAD diagram looks good.

Have you got a bearing supporting the shaft from the motor? Looks long, and a bit flexible under load maybe?

[Dessertrunner]

Motor drive shaft is still in development phase, I am thinking of using a tail shaft but even that has some issues. I will post photos tonight of were construction is up to. I want to find a flywheel for the motor so I can use the clutch plate, do you know were I can find a cheap fly wheel?
Tony

[Dessertrunner]

Project underway here is were its up to. Need to sort out drive from engine.
Tony

[Trevor]

Originally Posted by Dessertrunner

Project underway here is were its up to. Need to sort out drive from engine.
Tony

Looks good Tony,

Very interesting project and you are well on the way with it. Will you please explain the purpose of what appears to be a belt drive as shown in your photo.

Thanks, Trevor.

[Dessertrunner]

Hi Trevor,
The Telma retarder weighs in at 250 kgs and has a maxium rated speed of 4,000 rpm. In simple terms I want the dyno to handle a engine at 8,000 rpm so I have had to install a special belt drive to handle the job. For that reason I have put a 2.5:1 reduction with the belt. What is interesting is that at these revs chains won't handle it.
Tony

[Dessertrunner]

Adam do you bolt the engine solid or with the engine mounts?
Tony

[Trevor]

Originally Posted by Dessertrunner

Hi Trevor,
The Telma retarder weighs in at 250 kgs and has a maxium rated speed of 4,000 rpm. In simple terms I want the dyno to handle a engine at 8,000 rpm so I have had to install a special belt drive to handle the job. For that reason I have put a 2.5:1 reduction with the belt. What is interesting is that at these revs chains won't handle it.
Tony

Ok Tony.

Therefore the engine will be delivering 2.5 times more torque to the retarder than its actual output and this will have to be taken into account in respect of measurements. I presume that you have allowed for this when accessing the power handling capacity of the retarder.

For a given torque expressed in foot pounds and engine speed in RPM, the power can be calculated. using the standard equation -----

Horsepower = torque in foot pounds, times RPM, divided by 5252.

The constant 5252 comes from (33,000*ft·lbf/min)/(2π rad./rev.), because the degree of torque involves three measurements, i.e. applied force, the length of the applicable lever and the angle involved relative to time. Applying force via a lever obviously complicates things.

Cheers, Trevor.

[Dessertrunner]

Trevor,
I still need to get my head around the maths, what I can say is that retarder will handle 1,600nm continous load. which should more then cover it.
I get around the distance of the arm by calibrating it. The load cells output a linear 0 to 10v, we know that 0 kg so I use a know weight hang from the end of the arm to get the next point. From there I use the plc to do the rest.
The angle or position of the load cell needs to be in the same horizontal plane as the retarder centre shaft, as you say if that's not done I will be dealing with vector loads and that will get complicated.
Thanks for mention these point as I have been flat out trying to get my head around the drive shaft and not giveing these issues enough thought.
Tony

[dynomatt]

Looks good. Probably a stupid question, but I hope it will be bolted to the ground?

Flywheel from any EJ Series engine will work. Should be $50 or less.

M