Hydraulic Steering Ram Bypass Valve

[PeteW]

So I'm am putting together an Electric Hydraulic Autopilot drive that I can connect to the Autopilot Course Computer. So far I found a new 9" steering ram and a very questionable Octopus reversing hydraulic pump. Based on the capacity of the ram (290 cc) I'm going to need 1.2 L/min out of the pump. So I'm type 1 for sure.

I don't know much about hydraulics and I'm hung up on what to do for the bypass valve. This is the hydraulic solenoid valve that's normally open and connects to the clutch terminals on the autopilot computer.

In the open position the fluid in the ram is allowed to slosh back and forth allowing the wheel to turn freely. When energized the wheel locks up and the steering goes under control of the reversing hydraulic pump.

My options for the bypass valve are limited. There's the Raymarine 12V Solenoid Valve part# M81167 for $399.  There's the bypass manifold/solenoid that bolts to the Octopus pump. Have not found one of those yet. Then there is regular hydraulic stuff I can buy on the open market. The only OEM part I can find is called a Spool Type, 2 way, N.O. solenoid valve. (Hydraforce SV08-25) My issue with this part is it draws 1.4 amps when energized.  This valve is way bigger then I need at 6 GPM flow. And it's really blowing my power budget.

Anybody out there know hydraulics? This is a critical component for these types of autopilot systems.

Pete

[PeteW]

The bypass valve I choose to purchase for the electric hydraulic steering ram project is a solenoid driven spool valve made by WVO Designs ( Waste Vegetable Oil). New it cost $99.  Its rated at 3000 PSI in both directions. Unlike a poppet valve which blows out at 150 PSI in one direction and based on the number of complaints from sailors about their autopilot slipping under heavy load the spool valve seemed like the only way to prevent that from happening. At 12 volts it draws 2 amps which was a disappointment but I designed a current controller using $5 worth of parts that you can get from Radio Shack. The current controller allows full 2 amp current during "pull-in". 400 milliseconds later it falls off to a modest "hold-in" current of 800 milliamps.

Here is the complete electro hydraulic autopilot linear drive.

Compared to the $1800 a new system costs, my parts break down is as follows:
Octopus Type 1 1LPM Pump $100
12v Spool Valve  $99
Hynautic Hydraulic Ram 9" Stroke 290 cc. (new old stock) $450
Hydraulic hose and fittings $125
BMW brake fluid reservoir  $20

Next up I will mount this drive to the steering quadrant cross beam via 1/4" aluminum plate bracket. A Buck Algonquin split cap bronze tiller arm will mount to the rudder shaft below the Edson steering quadrant and above the 2" plain bearing in the cross beam.

Pete

[Della and Dave]

Have you thought about using a sequencing valve?  Logically, they are similar to a transister. Pressure in one line acts against a spring and operates a valve.  They can have manual overrides and can be normal open, or normally closed.  If you rigged it so that when the system powered up, it closes the bypass, no electrical energy required.  The only issue I can see is that if your computer ha a discreet output to disengage the autopilot, like if it goes open loop.  You might be able to just use that to shut off the pump, no hydraulic so revert to just pushing fluid back and forth. 

In a past life we used these to sequence the operation of big airplane landing gear so the gear unlocks before the gear doors open, before the gear hit them etc.  The logic diagrams look a lot like simple transistor logic. 

[PeteW]

I did momentary consider using what I believe the hydraulics industry calls a piloted valve. Where the pilot input operates the valve from either pneumatic of hydraulic pressure. Your analogy to the base input on a 3 terminal transistor is spot on.

While the autopilot is engaged the pump its either off ,turning CW or turning CCW.   Since the pressure in the hydraulic system may go to zero when the reversing hydraulic pump is momentarily off. I fear the piloted valve would disengage the "clutch" during that time. Maybe if there was an accumulator on the pilot input ?

The guy that could have solved this problem would have been my dear old dad. 35 years a hydraulic engineer for Grumman Aerospace. ( same company that owned Pearson). BTW,  If you need any landing gear parts for an F-14 Tomcat or an F4U Hellcat, I think I got that down in the barn.

Pete

[Della and Dave]

Wow, I have run into a few guys at the Reno Air Races that would give an arm and a good part of a leg for Hellcat parts.   Don't think there are any Tomcats in civilian use at the moment, but very cool airplanes. I can hear the Top Gun theme song in the back of my head   One of my extended relatives got carrier qualified in the hellcat and was flying out to join his first carrier when the war ended and was told to turn back.  He never actually did an operational carrier landing, or saw combat, but really loved the airplane. 

If I understand correctly, I wonder if two one way check valves and an accumulator plumbed with a controlled leak back to the resovoir would work?  That way CW would provide pressure one way, CCW would provide pressure to the accumulator the other, and the accumulator would hold the pilot valve open the rest of the time, as long as you didn't stay perfectly balanced in the null for too long.   It would mean a lag time between the autopilot going off line and the steering being free, which might not be good if it was too long. You could play with the orafice size on the leak back to fine tune it and minimize that problem.  You might be able to address that on the electrical side too.  Maybe some sort of logic that says "if autopilot engaged and pump off and accumulator pressure low, then power pump CW then CCW." Just to recharge the accumulator.  Essentially it would wake up and wiggle it's tail periodically in calm conditions, like motoring if there wasn't enough signal oscillation to keep the accumulator charged. 

The trouble with a purely electrical solution is the current draw.  From what I have read, it's an issue on many long distance boats as it is.  Most of the selonoid valves I looked at in McMaster Carr and Grainger were drawing an amp or two through the coil just to keep the valve in the open position against a spring.  Seems like a lot of power wasted. Maybe a shuttle type valve would be better from that standpoint.

There are valves where the selonoid just rotates the shaft on a ball valve, but then if you loose power, it remains in the position it was.  Maybe you could rig a relay that would discharge a capacitor to move the valve back if you lost power, that way you would only need power to move it and charge a capacitor, not to hold it open.  Not sure I like the failure modes on that though, could end up locking the steering if it failed

[PeteW]

The Electric Hydraulic Autopilot linear drive installation is nearly completed. In my installation I chose to mount the drive unit assembly to the steering quadrant beam below the Edson steering quadrant. I think this is preferred because it links the tourque between the tiller arm and the linear drive through the upper support bearing for the the rudder shaft.

The arrow points to the upper support bearing that prevents torsional movement of the rudder shaft. What you see there is the 2 1/4" dia. collar at the top of the 2" OD sleeve bearing.When I replaced the old beam I made it a pillow block design so I did not have to remove the Edson steering quadrant. I believe failure of the upper bearing support can lead to rudder post stuffing box failure.

The tiller arm shown is from Buck Algonquin. $175 complete bored and keyed for 1 1/2". To get it to fit below the steering quadrant I needed to lower the cross beam 1" to fit the tiller arm. There was plenty of extra key and keyway exposed so it bolted right on.


The drive assembly is mounted to 1/4" 12x14" piece of 7075. The additional bracket was added to remove flexing. You can test the strength of your installation engaging the drive clutch and applying pressure to the wheel. Nothing should move especially the rudder shaft itself when viewed through the deck plate opening.


Here's a view of the pillow block cutout in the quadrant stop beam. Not yet installed is the rudder position transducer and the mechanical quadrant stop blocks.

Incidentally I was able to reprogram my Autohelm ST4000 controller to be an ST5000 hydro, type 2 (hard) control head. Autohelm's one-size-fits-all design philosophy is way different than that of modern Raymarine's.

[Della and Dave]

Looks really nice Pete, and a fun project.  If auto helm knew you were doing this they would either void your warranty or try to hire you.  I'm curious about your mount, how does it pivot as the arm swings?  Is it a monoball in the pillow block bearing? I was also curious about why you chose 7075-651, in the aircraft world we have a love/hate relationship with it.  Great strength to weight ratio, but poor exfoliation corrosion resistance, especially in the short transverse direction.  If you have your mounting plate out fore some reason, you might want to Alodine and prime it for corrosion resistance.  When do you think you'll be able to give it a sea trial? 

[PeteW]

I'm not sure what became of the Autohelm trademark but I believe there was a honeymoon period that went on for a number of years after the leveraged buyout of Raytheon in the mid 1970's were Autoheln maintained some sort of autonomy as separate division. Beyond the online  documentation on legacy marine electronics, Raymarine doesn't seem to care what you do with those old designs. What I did learn is that when you configure an ST4000 to be an ST5000, you need to download a manual for the ST5000 to understand the new menu selections that don't appear in the ST4000 manual.  They built those units with all the code for multiple models already in there. Saved them a small fortune in foundry and production costs doing it that way.

I was impressed by the hardness of the 7075. As hard as steel. Super strength to weight ratio.  My aluminum guide gave it "Good" corrosion resistance. But it does have zinc and copper in it and it can't be welded.  There were better alloys I could have bought but  I choked when I saw the price.

The ST5000 really needs the rudder reference hooked up and calibrated. Then I will take it out and swing the compass. Although I am ultimately looking for a deal to buy a DX-10 Smart Pilot.

Pete

[Della and Dave]

Pete:  Good luck when you get to try it out.  At the risk of going too deep, the reason I made the comment on 7075, is that it's a bit of a hot button for me in my day job. It depends a lot on the heat treat condition and thickness, and the ratings can be misleading, especially in the marine environment.  T6 (T651 is the plate version) is the peak hardness and strength condition. 7075 is between 5.1 and 6.1% Zinc.  In the annealed condition, the zinc is evenly dissolved, so isn't much of an issue, but as it is heat treated, it comes out of solution, resulting in a lot of zinc in the grain boundaries.  This means you get a little internal galvanic party going on, which causes exfoliation as the zinc sacrifices to the aluminum, typically where an edge is exposed.  In your installation, you will see it long before it becomes a structural issue.  It's mostly academic if the cost is a higher or you can't find better alloys easily, or you need the extra strength.  I totally understand keeping it cheap and easy using stuff you can get.

This table, for anyone curious, is from the ASM (American Society of Materials) Aluminum and Aluminum Alloys Specialty Handbook.  It lists 7075 T651 as a "C" rating with the note that in thicker sections, it's "E"  The ratings go from A to E, if you can read the notes at the bottom of the table.  


I've seen "Good" as a rating before,  usually based on the ASTM G44 alternate immersion salt chamber test.  It's a good screening test, but it is fairly accelerated, only exposes for 2 weeks usually, so it doesn't show up exfoliation problems as well, usually just pitting and generalized attack. Good means that it is being attacked in the 2 week duration of the test, just not enough to severely compromise integrity, yet.  Grumman used 7075-T6 on the Grumman Goose spar caps and had problems in salt water operations.  

In the stress corrosion cracking tests, it rates "Low" which means samples stressed to within 50% of their ultimate failure stress failed in an alternate immersion test with 3.5% Salt water at a neutral pH within 30 days.  They use 3.5% because that is about what sea water is.  Search for a NASA study "MSFC-STD-3029" for more info if you are curious.  

If you are using stainless steel bolts, like SS 316, 7075-T651 looses to SS316 on the galvanic scale, so the bolt holes might be the first place to go because they have a galvanic driver, a confined area, and exposed end grain.  Again, it will take a lot of time and you will see it, but it might be worth a squirt of a water displacing corrosion prevention compound if painting is a pain because you have to take it apart. Sorry for probably more than you wanted, as I said, it's a hot button for me.  Hopefully someone will find it useful for other projects.  

I hope the sea trial goes well when you have it finished.  I usually find something isn't quite right when I have done projects like that. Looks like a really cool system.  We have a Raymarine wheel mounted Autopilot, but other than knowing it does power on, I haven't tried to use it yet.  There is something about the engagement lever that doesn't feel right, but I really don't know what it is supposed to feel like, so it might be fine.  The plan is to go to the boat this weekend, mover her to her summer home, start on our new battery installation, play with LED lights and start de-winterizing her, if it doesn't snow.  

[Risto and Liz]

Pete, I jumped over here from you other post. Thanks for the idea for putting the tiller arm below the quadrant. I don't even like the shaft lash when the quadrant stop hits the wood block, let alone with the force of an autopilot ram.
I am not too sure about the hydraulic ram since that is out side my experience. I might be open to giving it a shot if yours works well. If I decided to just bite the bullet and buy the Raymarine Type 1 linear drive do you think it could be mounted the same way? I like that it is pushing on the same structural member rather than another part of the boat. I don't get why they didn't use a more robust bearing at the cross beam. That plastics leave bushing seems inadequate. Do you know of a two piece pillow block that would stabilize the shaft and bolt to the underside of the crossbeam?

[P69]

Risto,

I couldn't find a pillow block, so I made one.

http://www.pearson365.com/forum/index.php?topic=1412.0

[PeteW]

Here's what the shaft bushing in the quadrant stop beam looks like.



Its plastic and unlikely that it will wear out or break. What wears out is the I.D. of the hole in the wooden beam. Mine rotted away to nothing because it was pine. To replace the beam I made a split block design that could be assemble in place without removing the aluminum quadrant. The hole in the beam was bored to fit the OD of the thrust bearing. The entire beam including the bushing hole were sealed with west systems expoxy.



Without a MIG welder set up for stainless, this is the best I could do.

If you want a new bushing you can get them from McMaster Carr. They come in a variety of materials such as plastic and bronze that will fit an 1.5" rudder shaft. And you can get either 1 3/4" OD or 2" OD. Naturally you will need to remove the quadrant to slip a new in place. Something I was trying to avoid.

As I think I mentioned the wobble in the design as-is from Pearson is because the beam lacks sufficient bracing to prevent side to side motion. Forward and aft it seems OK.

Pete

[Risto and Liz]

Thanks Pete. I think I can fab a solid, two piece pillow block from a piece of 1-1/2" Delrin by cutting a hole the size of the shaft, cutting it in half and bolting the two on the underside of the crossbeam. snug but not tight. We'll see.
A couple more questions about the autopilot setup. Do you think 3/4" 9 lam plywood could be used instead of the aluminum plate? Naturally, I would seal it all. It seems that It would be stout enough and might not flex as much.
Also, with the ram mounted on the crossbeam, where are the radial stop blocks?

[PeteW]

I like your Delrin solution. I'd cut it in half first, or simply buy two  pieces of rectangular stock. Bolt them together then have your machine shop bore the hole 1.5" -.0 + .005".  You will need to invest in an angle drill to transfer the mounting holes to your beam.

3/4" plywood flexes as does aluminum. In tension either is good. Under compression or shear they will both fail. So brace accordingly. A matching metal plate backing up the underside of the pivot mount will keep the holes from getting sloppy over time.

My tiller arm points forward.  I drilled the hole in the tiller arm based on the stroke of my hydraulic ram to be +/- 35 degrees.The stops (wooden blocks) were mounted to hit  slightly inside of that. Since I dropped the beam 1", I needed to fab taller wooden blocks for stops.  Although I don't think a hydraulic ram will be damaged in the same way electric screw drive type will be damaged. I need to post a final picture showing the rudder angle transducer and stops.

Even though Raymarine offers an electric hydraulic drive, I think Octopus makes the best one. Self contained fully compatible with Raymarine electric linear type I controller. I like that the pump rate (cc/min) is mechanically adjustable. You need to ensure that your drive will go stop to stop in 15 seconds or less.  No clutch to wear out or gears to strip. Up against a stop it will simply current limit your controller. They show up on EBay once and awhile. My homebrew drive has an Octopus pump mounted to the ram in a similar fashion using a pair of muffler clamps.

http://octopusdrives.com/products/linear-drive-0

I've been told by the Raymarine guy at West Marine that any non Raymarine stuff on the Seatalk back bone or connected in anyway electrically to their stuff in your boat will void your warranty. None of my equipment is under warranty anyway. So I don't care.

Pete

[Risto and Liz]

Thanks for the Octopus info. I called them and found them very helpful. I can get a good price through Fisheries Supply here in Seattle. I have the option of a 7" or 12" stroke length. It seems that the longer stroke would allow attachment further out on the tiller arm with better leverage and less lateral force on the shaft. Correct?
Do you have the part number for the bronze tiller arm? I don't see a two piece bolt on version.
I am going to use the Raymarine EV200 sail auto pilot. Pricey, but has some great features.