Small Engine Dyno: Designing & Attaching Torque Arm

[VitaX]

I am currently in the process of designing an engine dynamometer for a small engine. Maximum power of 1 KW @ 7000 RPM and a maximum torque of 1.6 Nm @ 5500 RPM. Here is my proposed method, but I have some concerns that I'll state below:

Prony Brake
Essentially a friction dyno, I would probably affix a disk brake to the engine drive shaft with a set of brake calipers to apply the braking load (friction force) to the shaft. My main question is, how would I attach the torque arm so that the braking load is thus transferred to the arm which presses downward on a load cell?

Here is an example of what I would like to follow:

What I'm having trouble seeing is how the torque arm is attached so that it will capture that braking load. In theory what I would need to do, is to apply friction the shaft via the brake pads, to some arbitrary RPM I would like to measure the torque and power at. Once I have the RPM and the load cell readout, calculating torque and power is simple.

Here is the standard prony brake setup:

What they are doing here is trying to balance the moment arm by adding weight as friction is applied to the shaft. Essentially the same principal as having the moment arm contact a load cell, only the moment arm would be 180 degrees apart and trying to rotate down onto the load sensor. I do like this setup, but I figured it was more practical for shafts that aren't rotating in the several thousands of RPMs. So I wanted to look into the brake caliper idea as seen in the first image. As you can see here though, as you tighten the nuts down, you are essentially creating a clamping force on the shaft thereby slowing the shaft down.

I hope I've made my question clear and have not confused anyone, if I still need to clarify, feel free to post and ask some more questions.

Main Question: I don't understand how they have the moment arm attached in the first picture so that the braking force that the caliper applies to the brake disk is transferred to the arm which then presses down onto the load cell at a given RPM. It looks to be able to pivot on the shaft. If anyone else has any alternatives to this idea, feel free to post them.

 

[Baluncore]

Brake dynamometers rely on a shaft rotating in fixed bearings with a brake rotor also mounted on the shaft, with a reaction arm, (centred on the shaft by a bearing), that prevents rotation of the brake stator. It is the scaled force from the long reaction arm that you measure.

Fundamentally, you must remove the 1 kW power so it does not cook your friction brake.
It would be good if you could recover the majority of the energy for utilisation or recycling.
It would also be good to have rapid and remote control of the “brake” setting.
The dynamometer you select will probably be dependent on the type of motor you are testing and the total test time expected.

You could consider a hydraulic system. That would use a small fixed displacement pump to move hydraulic fluid through an adjustable valve. The torque is the hydraulic pressure difference, the RPM is the flow rate. The oil then goes through a radiator or heat exchanger for energy recycling on it's way to the cool reservoir tank for the pump inlet. How can you use the heat, hot air or water produced?

If you are testing an electric motor then you should be able to use a vehicle heavy duty alternator as your brake by adjusting the field current. The RPM is the frequency of the alternator output. The output current and voltage give the power. You can correct for efficiency of the alternator by knowing it's internal resistance. You might then apply the recovered electrical energy back into the general supply.

 

[VitaX]

I've actually considered three other dynamomters (such as what you have listed). In terms of simplicity and efficiency for a low horsepower engine like the one I described in the OP, what would you suggest be the dynamometer? I only know of 4 possible choices: Inertia Dyno (Utilizing a flywheel as the load), water brake, AC electric motor/alternator, and the prony brake (friction dyno) as I described in the OP.

I had thought the friction brake and torque arm would be the easiest form of measuring the torque at varying RPMs. As for dissipating the heat on the brake pads, couldn't I just put a fan next to the brake? As you can see below, here is another representation of the prony brake and how they implemented their torque arm. To be honest, I don't like how the torque arm is situated like that, so I kept thinking of how I would change it to have a good braking system that was connected to my torque arm. Because essentially, doesn't the brake and torque arm need to be connected in some shape or form to transfer to the braking force from the brake pads to the load cell via the torque arm? The torque arm can't be situated on the shaft so that its a tight fit because then it would stall the engine. So as the torque arm is contacting the load cell, the drive shaft needs to be able to spin while under load. I'm just having an issue envisioning this, for some reason.

http://img822.imageshack.us/img822/8448/prony2.jpg

The Alternator idea would also be viable. I just wasn't completely sure how one would vary the magnetic field strength inside the alternator to load the engine in steps. What exactly would the wiring be like? I am actually intrigued by this idea because it has the least amount of mechanical moving parts, which means safer for people to be around. All I would have to do is couple the engine shaft and alternator via a chain and sprocket or flexible disk coupling, vary the magnetic field strength in the alternator to induce a load on the engine shaft, utilize the electricity generated to power something (perhaps a series of light bulbs), and read the current draw and voltage output to calculate torque and power. No torque arm or load cell required here.

The water brake is also a good idea, but I don't think they are typically utilized on such small engine power outputs like what I have.

The inertia dyno was another possibility. All I would need is just a flywheel to act as the load. Problem is, I haven't researched too much into this yet, so I don't know how one would measure the torque at a given RPM without data acquisition software.

 

[Baluncore]

Because essentially, doesn't the brake and torque arm need to be connected in some shape or form to transfer to the braking force from the brake pads to the load cell via the torque arm? The torque arm can't be situated on the shaft so that its a tight fit because then it would stall the engine. So as the torque arm is contacting the load cell, the drive shaft needs to be able to spin while under load.

...with a brake rotor also mounted on the shaft, with a reaction arm, (centred on the shaft by a bearing), that prevents rotation of the brake stator.

A bearing on the common shaft carries the reaction to the force on the load cell. If you are unhappy with a reaction arm on a pair of bearings at 7k RPM then how come you are not unhappy with the shaft mounting bearings?

A drum brake can be balanced about the shaft axis. A disk brake needs two calipers to avoid high side force on the shaft.

An alternative is to align the motor shaft with the reaction arm shaft. The brake then straddles the gap.

In terms of simplicity and efficiency for a low horsepower engine like the one I described in the OP, what would you suggest be the dynamometer?

The dynamometer you select will probably be dependent on the type of motor you are testing and the total test time expected.

An engine converts energy from one form to another. In the OP you have identified your engine as having a 1 kW shaft output.

But what type of engine is it? electric, internal combustion or maybe a turbine of some sort?
How long will you be testing it, for 10 seconds every 5 minutes, or maybe continuously for 24 hours?

 

[VitaX]

Its a single cylinder, 4-stroke internal combustion engine with a pull start. The test won't be ran very long, I would expect 5 minutes maximum if that. The goal of the friction brake dyno would be to start the engine, open the throttle all the way, and apply the brake until the desired RPM is reached in which you would like to measure the torque and horsepower. Apply the brake with more and more force to get more data points along the RPM spectrum, thus creating your horsepower and torque curves. I imagine that would take 5 minutes to 10 minutes maximum depending on how many data points you would want.

As far as the the mounting of the reaction arm, I don't think I fully understand what you are referring to with mounting the reaction arm on a pair of bearings. Could you explain this more in depth? I guess I'm having trouble understanding how one would mount the reaction arm to pick up the braking force applied to the shaft via the brake calipers without completely stalling the engine. Because as the torque arm places a load on the load cell, wouldn't that in essence stall the engine completely because you are stopping the drive shaft from turning? I was trying to picture from the first image I posted in the OP as to how the torque arm is mounted on the shaft to pick up the braking force and essentially transfer that force to the load cell, without completely stopping the rotation of the shaft in general.

 

[VitaX]

Sorry for the double post, but I think I understand what you are saying now with the torque arm mounted on the bearing itself.

http://www.smaltd.com/FreeWheels.htm

You are referring to this right? Where the bearing housing is essentially the load absorber and the reaction arm would rest solely on a load cell to absorb the load and give an accurate readout of the braking force applied to the engine shaft. Is this what you had in mind?

 

[Baluncore]

Yes.
The rotor of the brake always spins with the engine.
The stator of the brake remains attached to the reaction arm.
The reaction arm is pivoted on independent bearing(s), on the engine shaft, or aligned with it's axis.

 

[VitaX]

So essentially what I would have is the following:

An engine with a centrifugal clutch mounted on the flywheel
Machine a shaft of 3/8" diameter to mount inside the clutch housing
Along this shaft, I would have a set of ball bearings to support the shaft
The brake rotor would be mounted in between theses ball bearings on the shaft

This is where I'm trying to formulate how to actually measure the applied braking force on the shaft:

Pivoted on one of the ball bearings, is the torque arm that is also connected to the brake caliper (somehow)
As you apply braking force to the rotor, the brake pads and caliper should essentially try to rotate on that disk
Stopping that rotation is the torque arm which is pressing down on a load cell to give an accurate read out of the braking force applied at the corresponding RPM

Do you think that is how the reaction arm is setup in this image?
http://www.endless-sphere.com/forums/viewtopic.php?f=2&t=24024&p=348385&hilit=dyno

My main question is, does this mean we will essentially have the throttle at wide-open the entire test and just adjust the clamping pressure on the brake rotor to take our load measurements at different RPM?

Also, would an inertia dyno be more simplistic than this to make? I've read they are quite accurate for low horsepower engines like the one I have. From what I've read, you need a flywheel of known mass and diameter that you essentially couple with the engine drive shaft. We would need Dyno software for this one because I believe the software would calculate how long it takes for the engine to accelerate the flywheel and essentially plot both the torque and horsepower curves.

http://inertiadyno.com/

 

[Baluncore]

Yes, the main shaft there seems to run all the way through with a bearing between the reaction arm and the main shaft.

My guess is that a 3/8” shaft will be a bit weak for the expected maximum torque. Take a look at a similar capacity engine and see what diameter shaft it uses. My guess would be 5/8” or 3/4”.

In effect there is an input shaft and an output shaft with the reaction arm.
The friction brake partially couples those two shafts.
With sufficient support the one shaft can be cut so the bearings at the stator end do not have to spin.
The reaction arm can then be mounted in a low speed bushing but is still axial with the input shaft.

A disk brake will provide side forces on the reaction arm bearings. A drum brake will be better balanced and so not require as much alignment control.

Attached is how I would do it, using the rear drum brake assembly from a small car that is being scrapped. Direction of rotation is important for brake control so use the assembly drum from the correct side of the car. Attach the brake drum to the flywheel or shaft of the engine. Carefully centre and balance the drum on the engine. It is the only thing that moves fast. It needs only one plastic bush bearing at the end of a long tubular shaft that removes the requirement for exact alignment. It is designed for minimum machining. The existing flexible hydraulic brake line controls the brake.

If needed, a spring can provide an up-force on the shaft near the brake assembly to remove the brake drag due to brake mass when running free.