Hydraulic Dune Buggy Design: Is My Math Correct?

[larkinja]

Wondering if you have time for an offshoot question?

I think I am going to use an analog driver card to control the pump. I've been working with the card engineers to figure out the current ratings and such, its got the basics, adjustable ramping, separate adjustments for up and down. Adjustable dithering. Nothing fancy, but should work. It takes its input from a simple potentiometer, so this will give us huge flexibility in what we choose to use for a "throttle" device.

My question is about steering. We want to use a steering cylinder and a valve of some sort. Maybe 2 proportional valves, one for each direction, or some sort of bi-directional valve. Tonight I was looking through this device. Seems more complicated than it needs to be but it seems like the right idea

http://www.sauer-danfoss.com/stellent/groups/publications/documents/product_literature/520l0521.pdf

Do you have any experience with anything like this? I was hoping there would be some sort of basic analog device that would allow us to control the steering in a similar way as the pump control? I know I can use an orbital valve, but I don't want a traditional steering wheel. I know we can get a potentiometer joystick for under $100, so we may still play around with that idea if we can use a potentiometer for steering. Another thought if the joystick is to weird to drive is a yolk from an airplane. I took about 30 hours of private pilot lessons and the feel of the airplane controls is pretty cool. The thought would be that just a 90 degree turn to the left of the yolk would be a full wheel turn to the left, etc... Throttle could then be a thumb lever or a twist grip or a pedal on the floor with a pot box like what is used on a golf cart.

Anyway, the basic question is how can we steer with a potentiometer. And I mean fairly innexpensively, I realize there are some pretty amazing technologies that can be used, but we just don't want to spend a fortune here. We can always upgrade down the road.

The other thought is about reverse. When the pump reverses the flow to make the vehicle back up, seems the steering will be backwards, so I am guessing we would have to somehow reverse the steering direction when in reverse. Oh, for reverse, the card will have a couple relays that will reverse the polarity going into the valve. So putting the vehicle into reverse will require flipping a switch. I am thinking this switch could signal the steering to reverse as well maybe.

 

[drankin]

The main hydrostat pump will not provide oil for steering. This is never done because steering requires a dedicated oil source. You could oversize the make-up pump that is part of the hydrostat unit but even this is not a safe design for a steering circuit. Ideally, you want to piggy-back a small dedicated pump to the make-up pump.

Steering systems are a specialized dynamic application. There has been a lot of development in hydraulic steering to get it tuned and reliable.

The best and inexpensive way to steer would be with a traditional automotive power steering system. If we go custom hydrualic or electric over hydraulic we can use a joystick and performance will depend on how much you want to spend. A simple system will give you "bump" steering. You will bump the joystick in the direction you want to turn as opposed to holding the joystick in a turned position. Kind of like steering with buttons. How far you hold the joystick will only control how fast it turns not into what position it goes to. Not very natural. To have joystick angle position correlate with wheel turn angle would require an electro-proportional closed loop system. Very expensive. Or go with the Sauer valve systems but they are designed for a steering wheel and are probably more than you want to spend.

 

[larkinja]

The main hydrostat pump will not provide oil for steering. This is never done because steering requires a dedicated oil source. You could oversize the make-up pump that is part of the hydrostat unit but even this is not a safe design for a steering circuit. Ideally, you want to piggy-back a small dedicated pump to the make-up pump.

Steering systems are a specialized dynamic application. There has been a lot of development in hydraulic steering to get it tuned and reliable.

The best and inexpensive way to steer would be with a traditional automotive power steering system. If we go custom hydrualic or electric over hydraulic we can use a joystick and performance will depend on how much you want to spend. A simple system will give you "bump" steering. You will bump the joystick in the direction you want to turn as opposed to holding the joystick in a turned position. Kind of like steering with buttons. How far you hold the joystick will only control how fast it turns not into what position it goes to. Not very natural. To have joystick angle position correlate with wheel turn angle would require an electro-proportional closed loop system. Very expensive. Or go with the Sauer valve systems but they are designed for a steering wheel and are probably more than you want to spend.

Okay, you're right again. Okay, well I am familiar with using a steering valve, an automotive power steering pump and a double acting cyclinder for hydraulic steering. The thought crossed my mind that the larger the steering valves displacement is, the less number of turns it requires to move the cylinder a full stroke, right? Theoretically, if the displacement is high enough then would full steering be achieved with less that one rotation of the valve? Is there some way to calculate this? I think automotive power steering pumps generally create 2.5 or 3 gpm at around 900-1200 psi. This would already be on the engine, we haven't taken it off yet, so would be easy enough to use. I'm guessing this is probably going to be our best bet isn't it?

I like the idea of small movements on the steering wheel. Never having to take your hands off to make a full turn is the goal we're after. I was hoping a joystick would work, but it's looking less and less like that will work.

HAve any ideas? Or know a way to calculate the displacement to get the most travel from a small turn?

 

[drankin]

Okay, you're right again. Okay, well I am familiar with using a steering valve, an automotive power steering pump and a double acting cyclinder for hydraulic steering. The thought crossed my mind that the larger the steering valves displacement is, the less number of turns it requires to move the cylinder a full stroke, right? Theoretically, if the displacement is high enough then would full steering be achieved with less that one rotation of the valve? Is there some way to calculate this? I think automotive power steering pumps generally create 2.5 or 3 gpm at around 900-1200 psi. This would already be on the engine, we haven't taken it off yet, so would be easy enough to use. I'm guessing this is probably going to be our best bet isn't it?

I like the idea of small movements on the steering wheel. Never having to take your hands off to make a full turn is the goal we're after. I was hoping a joystick would work, but it's looking less and less like that will work.

HAve any ideas? Or know a way to calculate the displacement to get the most travel from a small turn?

I just don't have that much experience with steering circuits to help you there. If you could go with higher pressures and a smaller diameter steering cylinder then in theory you could turn more with less flow and still have adequate steering force.

 

[larkinja]

I just don't have that much experience with steering circuits to help you there. If you could go with higher pressures and a smaller diameter steering cylinder then in theory you could turn more with less flow and still have adequate steering force.

OK, no problem, I appreciate everything you are doing for us!

 

[ucom]

Very interesting thread. I did have a question, larkinja a few posts back suggested using a Bucher Hydrostatic Differential Lock Valve, I had an idea. Why couldn't you use it in conjunction with a set of solenoid valves to achieve a 4WD/2WD system. By opening or closing the valves with the brake pedal (much the same way your break lights come on) you could by pass the whole drive system and make it so you could use the disk brake idea. Also, in essences with the solenoid valves open it acts like a clutch in a normal car, with them closed the flow is diverted to the drives. In addition you could make the vehicle a 2WD (front wheel or rear wheel drive) or 4WD with the flick of a switch. I came up with a VERY simple layout sketch.

 

[larkinja]

Very interesting thread. I did have a question, larkinja a few posts back suggested using a Bucher Hydrostatic Differential Lock Valve, I had an idea. Why couldn't you use it in conjunction with a set of solenoid valves to achieve a 4WD/2WD system. By opening or closing the valves with the brake pedal (much the same way your break lights come on) you could by pass the whole drive system and make it so you could use the disk brake idea. Also, in essences with the solenoid valves open it acts like a clutch in a normal car, with them closed the flow is diverted to the drives. In addition you could make the vehicle a 2WD (front wheel or rear wheel drive) or 4WD with the flick of a switch. I came up with a VERY simple layout sketch.

The biggest problem so far is the flow rate. The pump is capable of at least 75gpm, and the valve can't handle that much. Actually I am having trouble even finding a gear divider that can handle that flow. The thought at one time was to gang 2 smaller dividers together to separate the front from the rear and handle the flow, then use the differencial lock valves at each axle. Although we're thinking the original plan might still be the best. That is if we can afford the large 4 port rotary divider.

 

[drankin]

The biggest problem so far is the flow rate. The pump is capable of at least 75gpm, and the valve can't handle that much. Actually I am having trouble even finding a gear divider that can handle that flow. The thought at one time was to gang 2 smaller dividers together to separate the front from the rear and handle the flow, then use the differencial lock valves at each axle. Although we're thinking the original plan might still be the best. That is if we can afford the large 4 port rotary divider.

What are you getting for pricing on the flow dividers? They shouldn't be very expensive. All a gear flow divider is 4 gear motors with a common shaft, a common port on one side and individual ports on the other.

 

[ucom]

What about this for a drive system. See Sketch. Remove the complex valve system all together and replace it with a simple high flow selector valve, and then use two motors with a mechanical link to provide the no slip differential on the two sides, this gets rid of the need for a high flow complex valving system and still gives you all the traction as before. The only draw back I see is that you would need to more drives in the system.

 

[larkinja]

What about this for a drive system. See Sketch. Remove the complex valve system all together and replace it with a simple high flow selector valve, and then use two motors with a mechanical link to provide the no slip differential on the two sides, this gets rid of the need for a high flow complex valving system and still gives you all the traction as before. The only draw back I see is that you would need to more drives in the system.

So, by drives you mean actually use 2 hydraulic gear motors with the shafts coupled together? Interesting idea. Is still think the flow rate is going to be a problem though. Even with the flow divided into 2 equal streams, that is still sending almost 40gpm into a single motor, and most motors have a 20-25gpm max. Correct me if I'm wrong, but basically that is what a rotary flow divider does isn't it?

 

[larkinja]

Drankin, I have a couple questions about the current schematic. One, the high speed circuit sounded like a great idea, but I'm wondering if it is necessary? The motors can only take 25gpm max, so diverting all the flow will just overspeed the motors anyway right? Can you think of a reason to keep it? Other than doing donuts, I can't think of any reason to have just 2 wheel drive anyway.

2nd question. I realized that in the text you have reverse function listed in the manifold. What does this mean. I thought we would be reversing the pumps swash plate and simply reversing the flow?

With this in mind, could we use the rotary flow divider as you have it, then just have a small solenoid valve to go between the two front motors, and one between the two rear motors? Seems like this would give us selectable lock on the rear, and a selectable lock on the front. With low flow going between the two motors, wouldn't this give us a limited slip effect without generating to much heat since it is just a lower flow balacing effect?


Let me know what you think.

 

[larkinja]

Another question. With the rotary flow divider, can the output from the pump have a large tee in it going to 2 2port rotary flow dividers? Seems it might be easier to find 2 port rotary flow dividers that can handle 40gpm that 1 4 port that can handle 80gpm.

Actually I am having tough time finding any the even come close in flow. Of the manufacturers that make them most seem to top out at 10 or 12 gpm. Do you have any suggestions for brands to look at?

 

[drankin]

Drankin, I have a couple questions about the current schematic. One, the high speed circuit sounded like a great idea, but I'm wondering if it is necessary? The motors can only take 25gpm max, so diverting all the flow will just overspeed the motors anyway right? Can you think of a reason to keep it? Other than doing donuts, I can't think of any reason to have just 2 wheel drive anyway.

2nd question. I realized that in the text you have reverse function listed in the manifold. What does this mean. I thought we would be reversing the pumps swash plate and simply reversing the flow?

With this in mind, could we use the rotary flow divider as you have it, then just have a small solenoid valve to go between the two front motors, and one between the two rear motors? Seems like this would give us selectable lock on the rear, and a selectable lock on the front. With low flow going between the two motors, wouldn't this give us a limited slip effect without generating to much heat since it is just a lower flow balacing effect?


Let me know what you think.

No prob, let's lose the hi-speed valves.

With the reverse function I have a series of "pilot to close check valves". The check valves are to protect the motors and flow divider from cavitation (self destruction). But in order to reverse you have to close those check valves so that the flow goes back through the motors instead of around them (and thru the flow divider going forward). I have it setup to protect the motors and flow divider in both forward and reverse. It looks big on the schematic but it will be a compact little manifold circuit.

As is shown in the schematic there is already an orifice between the flow divider motors for the front and rear sections. This provides the limited slip function. I could include this feature in the anti-cav manifold if needed. I based the schematic symbol off the "D series" Haldex rotary flow divider: http://www.haldex.com/Global/Hydraulics/Product%20Catalogs/flow_div_1205.pdf

 

[drankin]

Another question. With the rotary flow divider, can the output from the pump have a large tee in it going to 2 2port rotary flow dividers? Seems it might be easier to find 2 port rotary flow dividers that can handle 40gpm that 1 4 port that can handle 80gpm.

Actually I am having tough time finding any the even come close in flow. Of the manufacturers that make them most seem to top out at 10 or 12 gpm. Do you have any suggestions for brands to look at?

Using a tee is not a problem if that's how we have to go. We still need to add pressure filters on the two pump ports. Gear motors tend to create a lot of contamination as they wear and we will want to capture those particulates before they get to the pump.

I'll ask around for higher displacement flow dividers. If we don't find any that are reasonable we could go with a couple of two section dividers.

 

[larkinja]

No prob, let's lose the hi-speed valves.

With the reverse function I have a series of "pilot to close check valves". The check valves are to protect the motors and flow divider from cavitation (self destruction). But in order to reverse you have to close those check valves so that the flow goes back through the motors instead of around them (and thru the flow divider going forward). I have it setup to protect the motors and flow divider in both forward and reverse. It looks big on the schematic but it will be a compact little manifold circuit.

As is shown in the schematic there is already an orifice between the flow divider motors for the front and rear sections. This provides the limited slip function. I could include this feature in the anti-cav manifold if needed. I based the schematic symbol off the "D series" Haldex rotary flow divider: http://www.haldex.com/Global/Hydraulics/Product%20Catalogs/flow_div_1205.pdf

The Haldex product seems to be the closest to the right ratings, but looks like their biggest D series max flow is 14gpm per section. To get the max speed we planned, we need more like 20gpm per section. The pressure might be a bit low. I'm not sure. It says maximum inlet pressure 3000psi, maximum outlet pressure 4500 psi. I was planning the system to have a max of 4500psi as that is the max pressure of the motors. Not that we want to hit the max often. I can't say I totally understand how the outlet pressure can be higher than the in.

 

[RonL]

I can't say I totally understand how the outlet pressure can be higher than the in.

I think that can happen when a cylinder has what I would call a run away load, it acts as a pump and can pull a strong suction against the flow divider.

 

[drankin]

I can't say I totally understand how the outlet pressure can be higher than the in.

Actually, the outlet can be "intensified". Take a 2 section flow divider for example. If one of the sections has no load then you now have the inlet area of two motors rotating and providing a load to the outlet area of one section. This can effectively double the pressure output of that section if that much load is present.

 

[drankin]

Update: I've talking to an associate that works with hydrostats circuits often. He is going to pass on a circuit that will give you "posi" without a flow divider that kicks in when needed. I'll redraw the schem when I get the circuit.

 

[larkinja]

OK, sounds really good. Can't wait to see it!

 

[drankin]

I just received the circuit from him. It's pretty clever. He included a comment about your engine:

"When you mentioned that this project is using a 200hp engine, does this engine have a governor? I think it is very important to use an engine governor when running hydraulic pumps with varying loads."

 

[larkinja]

I just received the circuit from him. It's pretty clever. He included a comment about your engine:

"When you mentioned that this project is using a 200hp engine, does this engine have a governor? I think it is very important to use an engine governor when running hydraulic pumps with varying loads."

Great, I am excited to see it.

The engine does have a governor on it, and we can flash the computer to set that to whatever we want. Also, we are going to drive the pump with a re-drive, so we can make any ratio we want. ie when the engine is at 4000rpm, the pump could see 2000rpm, or whatever makes the most sense. We have to do some calculating to figure that out, but the peak horsepower of the engine is around 4000, so maybe we'll govern it to 4500, and have the pump see 3000 for a ratio of 1.5:1. Something like that. This should up the torque output as well.

 

[drankin]

larkinja, this is the circuit we've been looking for. It's a load-sensing circuit using logic elements that "talk" to each other and distributes the load across all drive motors. This keeps all drives rotating at the same speed when the circuit is activated (low speed, climbing, mud, etc.). And it can be de-activated when "positraction" is not required. I'll finish drawing the schematic tonight and post it.

 

[larkinja]

larkinja, this is the circuit we've been looking for. It's a load-sensing circuit using logic elements that "talk" to each other and distributes the load across all drive motors. This keeps all drives rotating at the same speed when the circuit is activated (low speed, climbing, mud, etc.). And it can be de-activated when "positraction" is not required. I'll finish drawing the schematic tonight and post it.

Sounds good, thank you!

 

[drankin]

Do you have a local fluid power shop that you are working with?

 

[larkinja]

Yes, a few of them. We ordered the motors from Applied Industrial Tech. I have also been working with Sunsource on some stuff.

 

[drankin]

Here is the latest schematic.

 

[larkinja]

Here is the latest schematic.

Sorry, I completely missed this post. Looks interesting. How would the controls for this work. Would it just be a switch that activates it. I'm not sure I understand how it would activate when needed. What type of valves would these be?

Thanks again for this, and sorry I didn't respond right away. I have been super busy with work and got a little distracted. We should be finishing the suspension this next week, and then we will begin mounting the engine and pump. The motors should come in in a week or so. We have the spindles made and the motors are ready to be mounted as soon as they come in. We already have the wheel hubs for the motors, so we could have a rolling chassis in a couple weeks. :)

Anyway, I would like to start researching the parts needed for your drawing, so let me know what I will need.

Will this be a cartridge system with a manifold, or just simply plumbing?

Thanks again.

 

[larkinja]

Also, in case this may be of use in your schematic. We are considering using a PLC to control the pump and the steering circuit, so there is a possibility that we could control this circuit with the PLC as well, once I understand how this circuit works. One of my buddies programs PLC's for a living and is very good at it, so hopefully will be an option.

 

[drankin]

Ah, you didn't disappear!

The schematic shows normally engaged but thinking about it it should be normally disengaged. All that you would do is energize that single solenoid to have all wheels locked in.

For normal operation you would run in standard parallel. In fact, since you have everything half built you could simply run all motors in parallel without the manifold as you were going to originally and see how that works for you. You would only engage the positraction at lower speeds. So you could test out the buggy at lower speeds and see how well it runs. If you find that one of the wheels takes off without the others a lot then we can finish up the posi manifold and add it as a performance mod. The manifold would take 6-8 weeks to be built and shipped.

It would be single all cartridge valve manifold that the pump and motors would connect to.

 

[drankin]

Also, in case this may be of use in your schematic. We are considering using a PLC to control the pump and the steering circuit, so there is a possibility that we could control this circuit with the PLC as well, once I understand how this circuit works. One of my buddies programs PLC's for a living and is very good at it, so hopefully will be an option.

Steering wheel or joystick? If steering wheel then I don't seen any reason to include steering into the PLC.