Q&A Forum
Jostein,
I asked some questions via email about selecting a hydraulic pump and valve for my e-bike dyno. The quality of your answers made me think that I should make this discussion public so others benefit from the time and thought you put into the answers. To bring everyone up to date....
After years of loyal service it is time for my homebrew power steering pump + oil cooler + bathroom scale style dyno need to retire.
I have attached a picture of the style of motor with which I work. These mid drive units are a blast. The are smaller and lighter than hub drive for given capacity and the keep the center of gravity down and in the center of the bike.
The motors are most common in a couple of ratings 250W, 750W, 1kW-3kW, and now 6kw. My current 3kW can easily hit a teeth rattling 40 mph.
As we identified in the email thread, testing an electric motor on a dyno has a few different challenges than testing a gas engine.
1. Dissipating the high end power is not a problem. 6kW is only about 8hp.
2 Dealing with low end torque is more of an issue. According to http://www.engineeringtoolbox.com/electrical-motors-hp-torque-rpm-d_1503.html a 7.5hp motor with a nominal speed of 3450 rpm will deliver 79 ft lbs of torque at 500 rpm.
3. The output shaft rpm can vary widely depending on how they are connected the the bottom bracket of the bike. The shaft speed of the motor pictured is rated at 5000 rpm. There 5:1 planetary reduction systems which reduces the unit output to 1000 rpm. There is additional 4:1 chain reduction. Some units have both the 5:1 and 4:1 reductions within the motor housing.
4. Rather than measure peak output. With an e-bike motor we want to measure the output torque, power, and efficiency as we vary the output RPMs. It was particularly interesting that your instrumentation package can control a servo which dynamically adjusts the pressure valve to keep the rpms constant.
Thanks again for this project. It brings accurate dyno test into price range garage builders can afford 🙂 I look forward to hearing more feedback on a pump selection for this application.
Hi,
Indeed an interesting application.
A brush-less DC motor has maximum torque at 0 RPM, and also maximum current at this point. So we need to find the torque, to make sure we don't over-drive the pump. Torque is what creates pressure, not HP.
The torque curve of a BLDC motor is pretty linear with RPM, so at 0 RPM the torque is max and at max RPM the torque is 0. In the middle of the RPM range you will have your max power. So if max power is at 1000 RPM and it is 6kW, you will have about 42lbft torque. Solving for 0 RPM you will get about 84lbft torque.
Pressure generated by the pump is given by the torque applied. The equation is:
P (psi) = 24 * pi * T (ft lbs) / Displacement (cu in/rev)
Assuming max pressure of 3000psi (check your pump spec) we get a minimum displacement of about 2 cu in/rev.
Note that in practice, the controller will need to limit the current at 0 RPM in order not to risk burning the motor. So I expect controllers will have some sort of soft start built in. So it is very likely you can go well below 2 cu in/rev without issue. Also, you may not be interested in the torque at 0 RPM, so you can start your run at say 50 or 100 RPM or so and that will enable you to choose a smaller pump.
Anyway here is an example of a quite inexpensive pump rated for 1.95cu in/rev:
Let me know if you need further input!
Cheers,
Jostein
Thanks for your thoughts and suggestions. I ordered the pump you list.
No one I have talked has any idea about the low end torque on these motors. There is a lot of marketing material available but little engineering data 🙁 The next step seems to build a prototype with just a pump, valve, pressure gauge, and a couple of thermistors to see if we are in the right ball park.
Thanks again. If everything goes well, you should hear back with some results in a couple of weeks.
David
As a quick update. I found a used Parker Hydraulic Flow Control Valve F1220S that will work well enough as a proof of concept.
The gearing was kind of kicking my butt 🙂 Rather than deal with the huge variety of output speeds on the motors I am researching. I decided to just make a chassis dyno. Found an old 7 in. quad v-belt pulley that I should be able to turn down to a drive roller that is suitable for bike tires.
Yes, I expect there is no absolute max torque that can be specified. For a short period a motor can generate much more torque than it can sustain for a long term. So you need to test and see what torque you can get while monitoring the temperature.
Cool application you are dealing with! Let me know if you want to order a kit, or if you want to test a bit first. It is about 2 weeks delivery time.
Cheers,
Jostein
I think I need to spend the next couple of week in prototyping to make sure the pump, valve, and drum work for my loads. Then I will take a week or so to figure out how to mount the drum and pump on bearings.
If all goes well, you will be hearing from me then.