Pressure Control Valve for ElectronicallyControlled Hydraulic Motorcycle Clutch

Even though you haven't described it fully enough for anyone to tell you anything specific about what components to use, etc., may I suggest that you use a hydraulic control valve and hydraulic actuators to control the clutch engagement. One side of the control valve can be a bypass port back to the accumulator, while the other can control the engagement. This simple arrangement does not allow for reversing the positive engagement, so if that's needed (I can only assume that it would be desirable), then two simple two-port servo control valves can be used rather than a more complicated single valve. Also, please note than PLC's aren't known for high-speed servo systems, so make sure it samples control signals rapidly enough to control your system.

Bill,

Thanks for your response. What sort of a hydraulic control valve would you suggest?

Here's the solenoid valve that I was referring to previously: http://www.summitracing.com/parts/SUM-760002/ It's just a 12v electrically operated solenoid that when voltage is applied prevents fluid flow. It just prevents the flow from returning to the master cylinder. I was hoping that by pulsing it I could get the desired results but that didn't happen.

I'm using a Galil RIO programmable logic controller to activate the solenoid. When I'm printing to the terminal each loop that does a sample, comparison and IO is taking 1-1.5ms which is adequate. Without the print to the terminal I suspect it should be quicker. The solenoid takes about 10ms to close ad 15ms to open.

The type of valve I refer to is a "hydraulic servo valve." See Wikipedia for schematic.

Some other links at Globalspec: Control Valves , Hydraulic Clutches

I've spent quite a bit of time looking at GlobalSpec for the right valve but of course I may have been looking in the wrong place. I need a combination of precision, the ability to work with low pressures (<250psi), low volumes (<20cc) and something that will take an analog input to control the pressure. If I could afford it that would be nice too.

For the time being, forget about pressures and flow rates. You need to characterize the system which needs to be controlled, and then develop a control strategy (philosophy).

If you're using solid state pressure sensors which are picking up spikes because of hydraulic hammer effects (which screws up the control), for example, then you may need to make pressure sensors immune to hydraulic hammer by adding dampers, accumulators, etc.

A lot of things we can do with electronics these days may have been controlled very nicely by mechanical means 100 years ago, so don't get to immersed in one philosophy that you don't consider other ways. It tends to clear one's mind, and occasionally, it leads one "outside the box."

I'm good with that. I've thrown one concept away and I can do it again.

I've been around machinery, computers and so forth for most of my life so I'm pretty good at coming up with ideas and understanding how things work. But trying to come up with a set of off the shelf parts to pull this thing together has been beyond me so far which is why I finally decided to ask for help.

You mentioned in another post that you want to control pressure, but hydraulic control systems typically operate via pump which has a maximum pressure and flow rate. In other words, the pump has a compliance pressure capability of say, 2,000 psi at 4 GPM.

With that hydraulic power supply, one may control motion of a hydraulic cylinder, or one may control the rate at which it moves, and with a little more effort and a controlled bypass, one may control the pressure. Since hydraulic fluid is more or less incompressible, and since most hydraulic pumps operate at constant speeds, flow is more or less constant. When resistance to motion is encountered, pressure starts increasing as a function of the resistance and it increases rapidly to the compliance pressure of the pump and its motor when a cylinder reaches the end of its stroke, for example.

Unless you are working against a spring-loaded clutch mechanism, I don't quite know how you may control a non-spring-loaded clutch actuation by controlling hydraulic pressure.

Are you looking to maintain a constant pressure are are you looking to vary the pressure?

I want to vary the pressure. I plan to write some software that will store a profile as an array of time/pressure settings and then upload the array to the PLC. Then when a triggering event occurs I would like the controller to start a cycle where it adjusts the line pressure every tenth of a second to a programmed value that it obtains from the array.

With a standard relief valve you could adapt a solenoid to apply pressure to the spring and ball assembly. So that varied voltage applied to the solenoid changes the pressure.

I had thought about something like that but just didn't know where to start.

Hey fellow enthusiast. As a racer I always try and keep things as simple as possible. I've run turbo bikes as well and looking back, fine tuning your own launching and driving skills and maintaining consistencies is always a challenge. With an adjustable hydraulic clutch releasing mechanism, it may allow you to have a bit more control even under the adrenalin while trying to feather your clutch under take off. As such I'd keep the wheel (mostly) down even while under full boost moments after launch.

I don't know if your a machinist or have access to one, but may I suggest a close variation from Bills suggestions in the form of a hydraulic servo system. It could be something simple to build while having adjustments for orifice size to and from the servo ports, Thus depending on design you can have some adjustments as you go through test and tune runs. Controlling fluid pressures in and back to the servos. They can be in the form of adjustment screws. Adjustments will be temperamental and incremental. It would also allow you to adjust the apply rate to the engagement mechanism. Using the hydraulic fluid for dampening while supplying fluid to the apply mechanism under varying forces, is basically what your trying to control while maintaining a linear flow of power to the wheels and timing all that with electronics. Very doable, but keeping it simple by design and function saves on some of the frustration of the outside variables.

Solenoids, particularly in this case, proved more outside controls as you've noticed and as Bill has mentioned with the hammering effect of the solenoid, spiking, etc, seems more complicated than need be. I assume other issues related to fluid densities and heat offer other variables that may effect the system you were thinking about. Are you using a wet or dry clutch system? There are significant differences when it comes to the type of oil used obviously in a wet system.

Trying to control engine torque under load to avoid wheel spin under full power is always a concern with short wheel based bikes. Again varying hydraulic apply forces under an adjustable time release may help.

Hope it helps some for what it's worth.

After thinking about it I remember making a variable control clutch release for one of my first turbo bikes. Looking back through some of my personal build books, looks like I had built one with a check ball and adjustable needle valve, one with a spring and servo with tensioner, and one incorporating a on/off noid. I've been looking through my stock pile of parts I've built to see if I might still have them. I'll probably build another anyway for one of my recent turbo projects that I may want to take to the track on occasion. If your at all interested I'll still keep looking. BTW, I also never used the clutch to shift after take off but controlled throttle and shift from the usual slight off power shifts. On one I made for a friend, we used a solenoid to redirect fluid back to the norm as he wanted to use the clutch after take off and kept his setting where he wanted it without having to mess with it further. You could really get fancy with this incorporating different release times per shift.

This is all quite interesting. I'd be interested in seeing how your system works.

My goal is to create a system that is instrumented and controlled to an extent so that I can tweak the clutch profile minutely and observe the results.

I've already discovered a few things about the behavior of the hydraulic clutch that I had not anticipated. One important discovery I made is that even when a solenoid is used to initiate the release of pressure (eliminating variations of the clutch release) the pressure profile is not smooth. In my case there is a hump in the pressure profile at about 45-50 psi. The centrifugal force imparted by the arms and weights makes things even more complicated. Then of course the desired clutch engagement itself is not linear. What I would desire is enough clutch in the first instant to move the bike as quickly possible, hopefully putting it right at the edge of lifting the front wheel, and then as the boost suddenly comes on compensating with more clutch in an effort to keep the motor off the rev limiter. I'm using a computerized boost controller and so the clutch engagement needs to be balanced very carefully against the amount of boost that the engine is being fed.

That's part of what's motivated me to try an approach that uses electronics - I can program any profile that I need to in order to compensate for these complicating factors.

My experience with the bike (which I have instrumented with a data acquisition system) is that my launch is not right in the first ten feet or so. I'm hoping that there is a significant amount of ET to be found in that ten feet by getting things perfect.

Maybe some additional inputs similar to traction control, wheel sensors?

or possibly a servo motor running a worm gear to release the clutch? seems like it would be easier to control the speed, & a opitical encoder for exact position control

just thinking out loud

This may not fully apply to your situation,but I remember reading about a motorcycle racer that used a star wheel mounted to the opposite end( side) of the shifter shaft to initiate a kill switch.He disengaged the clutch manually at the start, flipped a toggle switch(which enabled the micro switch on the star wheel).When he left the line, he released the clutch lever, and simply shifted without using the clutch at all.The micro switch unloaded the drivetrain as soon as the shift lever was depressed, and restored load when shifter was back to neutral position.(a fraction of a second later ).

The result was some unbelievably fast shift times, and no overheating of clutch during multiple runs.

Perhaps you could use this concept to initiate control of shift points.I would recommend a PLC with PID control and program each setpoint for each shift separately.Adjust the PID amplitudes for varying track conditions.

FWIW

HiTekRedNek

Well, again, I'm keeping it simple. With an incremental turn of a screw you can get the first ten feet of launch how you want it. That is always an issue with bikes. It will be more consistent in linear apply. It really is that simple. If you want incremental changes you can do so. You can even measure the fluids fill time and pressures. When you get those measurements when it's tweaked at the track, then possibly you can use that information to construct and program via electronics. The less into the programing and parts utilized the better. (even though my method is more analog) Our method has already worked in the past and worked well. Sounds like your more interested and bound and determined in making it a complicated application to electronics, which at some point you will succeed, good luck.

I have read the previous threads. Correct me if I'm wrong. your clutch is spring apply, hydraulic release so that the hydraulic actuator (cylinder) holds the clutch from engaging. You then wish to release the hydraulic oil in a controlled manner to "slip" the clutch in order to prevent excesive torque. If it is a hydraulic apply to engage there is different way to do this.

Simplist solution is to use a small (1/8" or 1/4") flow control (not needle valve!) to control the flow of oil inline with your solenoid valve. Varing the orifice controls the speed of the acuator. Unfortunatly it is one setting only, but stopping the solenoid may allow the acuator to hold at a certain apply or release position. a second solenoid in parrallel can be sequenced (thru PLC timing) for a fast apply/release.

This would be simpler than any PID, Servo etc. (plc would typically not be fast enough for rapid execution of PID or Servo)

Perhaps a variable displacement damper to absorb the hammer action from the solenoid.Sounds like what you really need is a PID controller for the hydraulic pressure.This allows adjustment of proportional, integral and derivative values which, when properly tuned, will give the results you need.I have a lot of experience with the old analog pressure controllers, which were very effective in controlling pressures, flows, temperatures, etc.I am sure there are now equivalent digital controllers that can accomplish the same thing.No need to reinvent the wheel, check for flow/pressure controllers with PID.This should be the answer.

Sorry it took so long to respond.Been off the grid and incognito for a while.

A side note:Automatic transmissions handle this problem nicely.Perhpas you could use the ATC from a car and the solenoids to get what you want.A honda, 1990 accord comes to mind.The solenod pack is mounted externally, on the outside of the transmission(2 bolts, 12 volt actuator via the AT computer), and the AT computer is in the passenger side floorboard.All components should be available from the salvage yard.It may not be easy to hack the computer for programming your needs, but not impossible.

Let us know how it comes out.

HTRN