Spool Valve Design 1

[LarryWoodcock]

We have some small (3/4&quot; thd) 4W 2P solenoid operated cartridge valves.&nbsp;&nbsp;These $54 valves have given us many problems and caused numerous headaches on a $12M manipulator.&nbsp;&nbsp;We intend to design and build our own spool and sleeve out of CRES materials and with non-stick hard coatings.&nbsp;&nbsp;Below are a few general questions and a more detailed explaination of what our problem is.<br><br>General Questions:<br>1. Does increasing spool clearance increase silt build-up due to increased by-pass?<br>2. What effect does increasing spool clearance have on spool shifting force requirments?<br>3. Are there any books that address the details of design of spool valves?<br>4. Do non-stick coatings reduce silting?<br>&nbsp;<br><br>Background:<br>We were having some intermentant valve failures early on in the project and could not pinpoint the cause (in hindsight it was most likely silting).&nbsp;&nbsp;The problems stopped when we put the system into integrated testing (different HPU).&nbsp;&nbsp;The problems occured again after the complete system was delivered for acceptance testing.&nbsp;&nbsp;The valves at this point had corrosion inside the sleeves.&nbsp;&nbsp;At this point several things were blamed for the valve failures.&nbsp;&nbsp;Bad proceedures that would allow air and thus corrosion into the system.&nbsp;&nbsp;Bad proceedures that allowed high temperatures to accelerate the corrosion.&nbsp;&nbsp;Deflections in the structure and improper installation torques that caused the valves to bind.&nbsp;&nbsp;The &quot;quick&quot; fix was to increase the spool clearance to make it more tolerant corrosion and deflections.&nbsp;&nbsp;During this fix we had a single valve fail again and made the determination that it was time dependent.&nbsp;&nbsp;This is were we learned the important silting lesson.&nbsp;&nbsp;By &quot;polishing&quot; the fluid we were able to extend time between failures.&nbsp;&nbsp;The valves have performed good for several months but are begining to have some problems.&nbsp;&nbsp;One valve was examined and corrosion was seen.&nbsp;&nbsp;Also part of the &quot;quick&quot; fix we &quot;tweaked&quot; the valves for flow and pull-in.&nbsp;&nbsp;We shortened the stroke and adjusted the spring to lower the required pull-in voltage.&nbsp;&nbsp;We also centered the spool around the ports to even the by-pass in each position.<br><br>The fluid we use is FDC 400 (a water soluble synthetic).&nbsp;&nbsp;The fluid is .5% water and is hydroscopic.&nbsp;&nbsp;The water content in the fluid varies and we sometimes see it as high as 3%.&nbsp;&nbsp;The water however is not free and does not cause any corrosion problems with other internal carbon steel parts.&nbsp;&nbsp;We do however see corrosion in the sleeve of the valve.&nbsp;&nbsp;&nbsp;The pitting is in the pressure area near the sealing lands.&nbsp;&nbsp;One assumption is that cavitation in this area is pulling the water out of solution and causing the corrosion.&nbsp;&nbsp;The cause of the corrosion is only academic.&nbsp;&nbsp;We plan on using CRES materials and coatings to avoid this.&nbsp;&nbsp;

 

[ghillis]

Building your own spool and sleeve seems like an expensive way to solve this problem.&nbsp;&nbsp;At $54 you are using some really inexpensive valves, and I am sure they are not considered suitable for your fluid by their manufacturer.&nbsp;&nbsp;Cartridge valves are geared more toward mobile hydraulics like trash trucks.&nbsp;&nbsp;My experience is that these systems seldom have good filtration.&nbsp;&nbsp;The saving grace is that they usually use oil based hydraulic fluid that makes things work better, and they aren't in extremely critical service.<br>You may be trying to cure the symptom, not the disease.&nbsp;&nbsp;The silt is coming from somewhere in your system (pump, hydraulic motor, reservoir, etc.), the clearances in the valve merely allow it to fall out of suspension and be detected.&nbsp;&nbsp;The polishing should reduce the wear rate since particles breed particles.&nbsp;&nbsp;I would look for the source by having the gunk analyzed.<br>Increasing clearance will make the valve more tolerant of particles since the fluid will have a higher velocity through the annulus around the spool it will carry more and bigger particles out of the area, and also increase leakage alot.&nbsp;&nbsp;If the valve is properly designed and the fluid is clean, there is a tiny reduction in forces by having more clearance.<br>Non-stick coatings are normally used to prevent corrosion in water based fluids and to differentiate products (sales tool).&nbsp;&nbsp;I think you need them on both spool and sleeve to derive any benefit.<br>What appears to be corrosion may be cavitation damage.&nbsp;&nbsp;Check the rated flow capacity of the valve against the actual, it may be undersized.&nbsp;&nbsp;The book I reference has some good info on cavitation.&nbsp;&nbsp;If localized just downstream of the flow restriction, its cavitation.&nbsp;&nbsp;CRES is more resistant to such damage that carbon steel.&nbsp;&nbsp;Moisture in the fluid probably makes the problem a lot worse, dessicant breathers on the reservoir prevent ingress of moisture.<br>High pressure filter(s) downstream of the pump is probably the easiest fix.&nbsp;&nbsp;This is one of the few places one can normally insert something into a cartridge manifold system.&nbsp;&nbsp;If this is a high duty cycle system, and you have some room, a kidney loop filter system could fix the problem.&nbsp;&nbsp;This consists of a relatively low pressure pump that circulates oil from your reservoir through a good (fullers earth) filter, a timer can control how frequently it runs.<br>Book - Lyon's Valve Designer's Handbook -Van Nostrand and Reinhold

 

[LarryWoodcock]

Thanks for the input, I will look for the book.&nbsp;&nbsp;Here are some comments and clarifications.<br><br>The expense of designing and building a valve not an issue if we can produce a reliable valve.<br><br>You may be trying to cure the symptom, not the disease...<br><br>Maybe, these valves and the corrosion in them are probably a source of the contamination.&nbsp;&nbsp;We are currently running 2 micron filters on the return and on the pressure side.&nbsp;&nbsp;All the dynamic seals in the system are teflon based (both carbon and glass re-inforced).&nbsp;&nbsp;All elastomers are Viton.&nbsp;&nbsp;All the joints are rotary actuators (similar to a vane actuator designed in house)and we have not seen any signs of the hardparts wearing. It is mostly Titanium with a little 17-4.&nbsp;&nbsp;All seal surfaces are TiN or TiCN coated. Our cleanliness usually falls between 14/11 and 16/12 (measured at the supply going to the arm).&nbsp;&nbsp;Maybe we should check what is coming back from the arm and or have the filters analyzed.<br><br>You said that you think that there should be a tiny reduction in forces with increased clearance.&nbsp;&nbsp;That was my thoughts going into this thing, but it seems that there may be significant increases in forces due to the increase leakage.&nbsp;&nbsp;I would attribute it to the higher velocities and thus increased drag. BTW our leakage is close to .1 GPM in each position.&nbsp;&nbsp;I have to test some valves for spares today I'll look at this more closely.<br><br>We are pretty sure that the corrosion is corrosion. It is upstream of the land.<br><br>Our HPU is closed (ParVent bladder type accumulator in place of the breather cap).<br><br>We have a kidney loop on the system for circulating the fluid through a heat exchanger.&nbsp;&nbsp;We have a separate heat exchanger on the case drain of the pump and the kidney loop never kicks on.&nbsp;&nbsp;Maybe we will convert it to a filter loop.<br><br>On materials, we may go with a ceramic spool and sleeve and reduce the clearance.&nbsp;&nbsp;Any thoughts on this?<br><br>FYI to see a few pictures of the manipulator in service go to <A HREF="

http://www.spacehab.com/ssrms/"

TARGET="_new">

http://www.spacehab.com/ssrms/</A><br>

 

[ghillis]

That's pretty cool, it is a little more sophisticated than I anticipated (CRES should have been a hint).&nbsp;&nbsp;<br><br>Your cleanliness seems OK.&nbsp;&nbsp;I'm having a hard time picturing your hydraulic circuit, a 4 way 2 position valve is going to run your manipulator lock to lock (fully extended or retracted). Are you using PWM or a proportional control on these valves?&nbsp;&nbsp;High cycling could create alot of wear.<br><br>Upstream of the land in the pressure or return flow? I'm afraid that I would need to see a cutaway of the valve to help you out.<br><br>Your source of water may be that the seals are designed to retain pressure rather than exclude it.&nbsp;&nbsp;Whenever external forces are applied to the vane, one side is pressurized and one side draws a vacuum, pulling in water.&nbsp;&nbsp;A significant positive pressure in the system is required to overcome this.&nbsp;&nbsp;Sampling your return should indicate if this is a problem.<br><br>I agree that the flow drag can be significant, but I assumed that the leakage across the lands would be symetrical, but this being a cartridge valve that would not necessarily be the case.<br><br>I have only had experience with flat lapped plate type ceramic valves, Marstan makes some (used to, anyway) that are good for high pressure, but on the heavy side for space flight.&nbsp;&nbsp;The neat thing about them was that the plates were spring loaded and would ride over particles instead of having to plow through them.

 

[LarryWoodcock]

These valve do not control the motion.&nbsp;&nbsp;That is controled with Atchley servo valves.&nbsp;&nbsp;The actuators are vane actuators not motors.&nbsp;&nbsp;The valves in question retract and extend vanes to allow unlimited rotation.&nbsp;&nbsp;There is some PWM on these valves but only to limit the heat build up by lowering the voltage after the solenoid has pulled in (about 50% duty cycle).<br><br>I don't understand your comment on the vacuum.&nbsp;&nbsp;We always positive pressures inside the hydraulics.&nbsp;&nbsp;The fluid is slightly heavier than water therfore always at a greater pressure.<br><br>Wish I could give more details about the joint but it is proprietary at this point.

 

[ghillis]

The pressure is greater statically, but dynamically, the pressure <i>can</i> be lower.&nbsp;&nbsp;Anytime there is a pressure surge, there is a corresponding rarifaction in the reflected pressure wave.&nbsp;&nbsp;Atchley is a good valve, but it should be the valve that is more sensitive to contamination.&nbsp;&nbsp;<br><br>Your application of PWM is on track.&nbsp;&nbsp;Unless the solenoids buzz, I would not think that's the issue.&nbsp;&nbsp;I'm stumped.<br><br>