600 kW / 850Nm hydraulic pump - need advice

[MortenHansen]

Hello all fluid power professionals!

We are in the early planning stage of designing an hydraulic engine dynamometer for internal use. We are looking at using a hydraulic pump as the load for the engine. We are planning to apply load to the hydraulic pump with a servo valve partially obstructing the flow of the hydraulic fluid. To measure power, we can either measure the pressure of the fluid and multiply with the RPM and a correction factor or measure the torque with a strain gauge and multiply with the rpm.

We need some advice on where to find a reasonably priced hydraulic pump that can transmit approx 600 kw / 850 Nm thru its shaft reliably.

The RPM of the engine is max 8500 RPM.
We might have to gear down this to a RPM the hydraulic pump can handle.

I was thinking about hydraulic pumps for dumper trucks and digging equipment like CAT or Komatsu. This kind of equipment is mass produced and I would think a lot of spare parts suppliers exist.

Also, where can we find servovalves that can handle this pressure with a fast response ?

We would like to cool the fluid via a water to oil heat exchanger. Any hints on suppliers ?

This has actually been done before in an chassis dyno application for cars.

The unit in left front of the picture looks like the hydraulic pump they use (does anybody on this forum recognize the brand of this pump by any chance?).

http://www.flyinmiata.com/images/misc_images/rototest/DSCN0238.jpg

Looks like the unit on the right with the piping is the servovalve unit that applies load to the hydraulic pump. The black unit on top is probably the servomotor for controlling the servovalve. A pc is attached that runs some real time control software.

The brake assembly is attached directly to the driveshaft of the car.
I think there is a gearbox attached between the driveshaft and the hydraulic pump. Anyone that know a manufacturer compact of gearboxes that can handle this much power and torque (600kW / 850Nm) ?

Here is a link to a pic of the servovalve they use

http://www.flyinmiata.com/images/misc_images/rototest/DSCN0237.jpg

Anybody recognize any of the components or is it custom?

We are not going to attach our dynamometer to the driveshafts, but directly to the engine crankshaft thru an axle. Therefore we need it to be able to handle at least double the power of this unit since the car differential distrubutes power to two driveshafts and two units.

Regards
Morten

 

[MortenHansen]

Alternative solution (question - what hydraulic pump can we use for this alternative) :

What about if we wanted to take the power straight from two wheel axles instead of one and run it thru the normal gearbox and shafts / differential of the car and reduce our demand for power to approx 550 HP (412 kW). Distrubuted over two axles this should be approx 205 kW per pump?

With 16" diamater wheels the wheel rolling distance per revolution would be (15"/2)*Pi*2 = 47 inch / rev.
This translates to 1,19 meter per revolution.

If the max test speed would be set to 300km/h then the rolling speed of the wheel surface would be maximally approx 5000 meters / minute. This would mean 5000(m/min) /1,19(m/rev) = 4200 rev/min max pump RPM. Then we would probably have to use a 2:1 reducing gearbox.

It looks like this is what is done here :

http://www.flyinmiata.com/images/misc_images/rototest/DSCN0238.jpg

When I look closer at the pic. it looks like there are TWO pumps in series on this application. It appears to be one valve control block at the output of one of the pumps (the inner) and one pipe feeding the valve block from the outer pump.

There are two hydraulic lines going to the right. -Looks like those are return lines that go to the cooler. Perhaps there is a fixed orifice in the small block on the body of the front pump that feeds some of the output from the front pump to the control block and the rest to the cooler.

What do you think ?

regards
Morten

 

[MortenHansen]

Btw I meant two pumps in paralell !

Regards
Morten

 

[MortenHansen]

Hallo Waross, I will check that out. Thank you.
However I have the feeling that a genrator that powerfull will be very harvy. The equipment must be movable by hand.
What do you think ?

Regards
Morten

 

[Skogsgurra]

Morten,

I think that a hydraulic solution will be in the $ 50,000 - 100,000 range. And you will have to deal with lots of waste energy. Most car and truck manufacturers use generators and inverters instead. The control is very good and the accuracy also. And the energy can be fed back to the mains - you can actually get paid for it instead of paying for getting rid of it. If you are in a big enough factory, the energy doesn't have to be fed back to the utility but can be used by your other equipment.

budt, ScottyUK, itsmoked and waross have all mentioned this possibility and I second that. A dedicated brake from Siemens with all its measurement and automation gear will surely cost a lot. But rolling your own motor/inverter will not cost much. I can guarantee that the all electric solution is cheaper than the hydraulic one. And you get rid of the heat - perhaps even get paid for it.

Gunnar Englund

http://www.gke.org

 

[MortenHansen]

Hello Gunnar, thank you for your answer.

I don't think feeding energy back to the net is so easy as it sounds. It is AC and the generator will not be in sync with the net at all. The engine is the "master" The output had to be rectified and then switched with the correct 50 Hz frequency. Switching 600 KW will take a VERY big stack of IGBT transistors and possibly some hefty transformers and capacitors as an energy storage for storing the exess energy until you can feed it in phase into a cycle. Also I think the genrator(s) will be very heavy.

I must admit that I think that paralell connecting a number smaller hydraulic pumps with a control valve will be the lightest and most portable unit. The unit will be run for only minutes at a time and only intermittently and there will be no connection for us to the net at all locations that could handle 600 (!) kw. This should be a portable system that can operate from 220V 10A auxiliary supply for the neccessary electronics. The hydraulics will have to do the work.

I know this has been done with hydraulics as the pictures show. We want to make a similar system to this or a bigger one with just one axle.

I start to think that separating the load on two axles will be the best.

Regards
Morten

 

[Skogsgurra]

OK, I can understand that energy is not an issue in your case. And that a hydraulic solution is better suited because of the portability needs. And if there isn't a grid stable enough to receive the energy - it just can't be done.

But you really need to update yourself on the electronics side! I see dynos that do what you want to do more or less daily in the Volvo, Saab and Scania development labs. And I do some commissioning on them as well.

A big dyno with something like 1200 HP braking power is not very big. And the motor isn't either. About the same size, or smaller, as the diesel engine it is coupled to. And the rectifying/synchronizing part is taken care of by standard inverters. Your concern about different frquencies and synchronizing to the grid is understandable - but completely unnecessary. It is being done in thousands of installations day and night. All around the world.

Gunnar Englund

http://www.gke.org

 

[budt]

Morten;

Another far out thought:

Get a propeller from a DC-6 that has variable pitch for your resistance. A prop that size could be chopped to reduce its diameter since they were driven by over a 1,000 HP if I remember correctly.

The variable pitch would give some latitude on torque also.

Did you find the Tyrone site. I believe they are owned by Parker now Here is a site that also makes them:

http://www.perfhyd.com/products/hotdeals/TP6700121001.htm

They make single and double pumps that could be big enough for your setup.


Bud Trinkel CFPE
HYDRA-PNEU CONSULTING, INC.
fluidpower1 @ hotmail.com

http://www.fluidpower1.us

 

[unclesyd]

Have you checked this one out.

http://www.taylordyno.com/eng/maineng.html

You might also check out this one.

http://www.hbus.haldex.com/products/fds.htm

The hydraulic dynamometer that you are looking at doesn't use the hydraulic power to really measure the output, it uses a load cell setup with their software for the power measurements and dissipates the hydraulic power and heat through the fans and cooler. Their literature also states that they use a high speed servo valve (patented by them).

 

[waross]

Re one or two units; I wouldn't trust the spider gears to last for very long if you locked one wheel and ran your power through the differential to the other wheel.
I agree with the figure that someone posted that your heat output at 600 kw will be over 34000 BTUs per minute. That's enough heat to boil about 34 Lbs of water a minute. (15.5 liters).
If cost is an important consideration, consider using the brake drums from a piece of heavy equipment and water cooling them. The technique was used in the logging industry for years. It's a good idea to start the water sprays on the outside of the brake drums before the heat builds up. Boil the water off and vent the steam.
Re portable; That is a challenge. Anything that has enough mass to stay in place and stay together while disipating 34,000 BTUs a minute will be too big to pick up. One man can move a very large load by himself with a pallet jack. If you have to move from site to site I suggest a dedicated truck with a small crane to load, unload and place the dynos, whatever method you choose to disipate the energy. Another old but cheap and dependable device is a water cooled (By sacrificial vaporization if necessary) eddy current brake.
respectfully

 

[MortenHansen]

Waross, there is a real time control unit that adjusts the load on each of the sides so that the spider gear in the differential is not seeing any different RPM. This way the spider gear will not be stressed.

I have actually found out now that the system is cooled via some fans that are driven by hydraulic motors. There is one 30 HP hydraulic motor on each fan ! This will cool the system down. I spoke to somebody that has this system in use.

If anybody knows if this is a piston pump or another type of pump, that would help. It doesnt look like a piston pump to me.

regards
Morten

 

[zeusfaber]

Another hydraulic approach might be to use a variable displacement pump driving against a fixed restriction.

Some of the cost of going to a more expensive pump can be offset by the fact that if you want to use a servovalve to control the pump stroke (to maintain required rpm), the valve need only be pocket sized.

You've still got to lose all that energy somewhere - if your duty cycle is low enough, the answer to this might simply be to use a much larger tank.

A.

 

[MortenHansen]

Hello Zeusfaber, thank you for your input.
That might very well be a good idea. I dont know how a variable displacement pump works. Does the stroke of a piston pump change by the control input or something like that ?

However we have also found a very low cost way of making a very fast servo valve. The response of the system must be very fast. This is neccessary to be able to control the rpm of the load load. The load can have a very high torque gradient in some RPM ranges. Therefore we need to apply load very fast. I dont know what loop response you will get with a variable displacement pump ?

Regards
Morten Hansen

 

[zeusfaber]

Morten,

Yes, a variable displacement pump works by varying piston stroke.

A typical axial piston design will have seven or nine cylinders in a spinning barrel (like the chambers in a six shooter. The ends of the pistons stick out the bottom of the barrel and engage with a swashplate. When the plate is perpendicular to the spin axis of the barrel, the pistons don't stroke at all. As soon as you tilt the plate, the pistons stroke up and down the cylinders with the rotation of the barrel. The other end of the barrel mates with some fixed ports and valve gear to connect each cylinder to flow and return at the appropriate stages of their journey.

The tilt on the swashplate is often controlled by a small hydraulic servo system, and can be very responsive.

Details of one model [link ftp://www.denisonhydraulics.com/Literature/AXIAL%20PISTON%20PUMPS%20&%20MOTORS/SERVICE%20INFORMATION/Premier%20Series/P16-260/S1-AM022-B%20P16C-260C-SRV-Complete-01-84.pdf] on the Denison website [/url] (notes: Document is 6 MB+ and link is ftp. This particular pump looks not far off the size you'd need, and appears to promise a 10% to 90% flow response of under 100 ms.

Downside of these pumps is that they are never really "low cost"

A.

 

[MortenHansen]

Hello Zeusfaber.
That is interesting. However we might need faster response than 100 ms. I havent done 100% of the loop calculations yet but we will definately take a look at that unit there.
Thank you again.

Regards
Morten