Peter and Bud - Upset Circuit

[TimWilborne 108]

Sorry, these are the best picures I got before I took it apart Below is the upset manifold. Notice on the right is the D10, then it goes through a pressure reducing valve, then to the servo valve. Notice the piping coming out the left, then down and over. You can see the T below the D10, then you can see it connect to one of the ports to the left of the accumulator. This is the best picture I have of the rest of the piping. Pretty much it goes down, over, and back up. There are holes in the frame where it used to go into the bottom side of the cylinders but they must have put them in with the machine upside down. Also it is kind of hidden but there are two upset cylinders. One is right b behind the angle coming through the machine in the picture. Look closely and you can see the piping

[gravitar 3]

Musta missed something.. Why are they upset? Pretty cute that you name your cylinders.. "Peter and Bud" :) Hope you can cheer them up!

[Peter Nachtwey 15]

I don't know why I should be upset. I don't even know what an upset circuit is! I think TWControls will tell us though. Bud, were famous! We have cylinders named after us! :) Seriously, I think TWControls is just trying to get some ideas about re-arranging his hydraulic system. As long as he keeps the tubing between the valve and cylinder as shout and solid as possible I think that is OK. TWControls already as a very short and and large diameter cylinder and that is good. He can get by with a little tubing but no more than is necessary.

[TimWilborne 108]

Do you think I should stick with the schedule 80 pipe? Very funny about my cylinder naming. I believe they are called upset cylinders because they forge the molecular bond by slamming the steel together but I have a technical definition at work. Let me look it up

[TimWilborne 108]

Ok, it is named after the Upset process of the machine which is performed by these cylinders. During this process it takes two molten pieces of steel and slams them together which molecularly bonds the two ends. I can get more in depth with this process but I don't want to put everyone to sleep. Pretty much is upset the molecules Now these cylinders also perform the Burnoff, Preheat, Flash, Trim, and Post Heat processes. Burnoff Cylinders - Sounds ok Preheat Cylinders - Boring Flash Cylinders - Cool sounding, but when using it for true butt welding this process is not used Upset Cylinders - Cool Trim Cylinders - Boring Post Heat Cylinders - Boring That's the best explanation I can come up with Since my cylinders are apparently now named Peter and Bud, the one on the left has given us quite a fit in the past. Would that be Peter or Bud?

[TimWilborne 108]

Upset Circuit Schematic

[Peter Nachtwey 15]

What is valve 19 and all that stuff in parallel? It must be your pressure/force control. Hopefully it isn't active when trying to control the two cylinders with the servo card. What is valve 31 for? It looks backwards. I don't see how it helps. Maybe Bud knows. Sounds like a mechanical problem, call it Bud.

[BudMan 1]

Wow!! first they nam a beer after me, then small plastic enclosures and now cylinders. I am flattered!!!

[BobLfoot 302]

TW I kinduh thought that upset had to do with forging, but we did ours with a flywheel and ram. An seperate induction heater took the billets to 2250 and then they were placed into the die before being pressed. 25 strokes per minute and several 1000 tons of force appled each stroke. Turned a 2.5 inch diameter shaft into a 2.5 inch diameter shaft and a plate of 10 inch diameter for an axle on a truck.

[TimWilborne 108]

19 is our servo valve. It is a HR Textron 27E50F-4A06-999. 26 is a pressure reducing valve for the servo 31 is a flow control to regulate the fluid going to 17 which is the D10 that is paralleled to the servo valve. The servo valve controls the cylinders during all phases except the upset process. During these times the D10 is not used During the upset the steel must be pushed together very quickly. At this point the servo shuts off and the big valve shifts to slam the steel together. The valve is then shutoff before it actually hits the upset distance. By the time the valve actually closes and settles, it will be at the upset distance. When the valve is being closed the servo takes back control of the cylinders. During the time that the servo is not controlling the cylinders flow control 31 is the only method you have of controlling the speed of the cylinders. Too slow and the steel will cool too much. Too fast and you will over shoot the upset point and press cold steel against cold steel. Either way it would be a bad weld. This is the most critical point of the process. It takes about .250 seconds max Seems like a crude way of doing it but new machines both larger and smaller do it this way. I really don't know why unless large servo valves are uncommon or would not have the precision necessary And you are correct. The left cylinder would be named Bud. The nuts that hold the rod in place were awful about backing off.

[TimWilborne 108]

Yes I have heard of these but have never actually seen on. The principle is the same except instead of induction heating it moves the steel together very slowly causing it to arc and heat up

[geniusintraining 2]

gravitar, To me its obvious why they are upset...if I were that dirty I would be upset as well , cleanliness is always better for hydraulics.... Sorry TW...had too

[TimWilborne 108]

I can make it clean, and hopefully with a little guidance from you guys I can make it so it will stay clean The Upset Cylinders new (clean) home

[geniusintraining 2]

nice....Happy hydraulics

[TConnolly 13]

Just for the hell of it, I colored in your hydraulics print. The only thing I was not sure of is if the directional control valve and the servo valve are ever both on at the same time, so I didn't color that part. This was a quick and dirty coloring job, so I might have missed something. Your servo valve symbol shows two torque motors for spool actuators, there is usually only one which diverts flow to either side of the large spool, but I presume it was drawn this way for simplicity (we usually show the pilot in our drawings), anyways, I digress. So as far as the green indicators are concerned on the servo vavle, they don't really represent an actuated solenoid, they show which side of the spool is active. We usually provide our maintenance staff with colored drawings showing pressure and return paths and pressure gage tables for each state of the machine. Edited 22 Aug 2006 by Alaric

[TimWilborne 108]

The big valve and the servo are never on at the same time The servo valve is not two separate motors and not a pilot valve. The schematic is a bad representation. It is the round cylinder that is on the left of the first picture. It is only capable of flowing a small amount of fluid to the cylinders. I guess in larger applications there would be a pilot valve for the servo. The diagrams. Now that is a great idea. Especially on applications where two valves and practically two different circuits can provide oil to the same cylinder. I am going to do the same and incorporate the diagram into pictures on the Panelview. This way at a particular point in the process, maintenance will be able to see where the fluid should be coming from Great Idea. Thanks!!! TW

[TConnolly 13]

Thats what I kind of figured for the servo valve, I've never seen a servo with two torque motors myself. I also put the diagrams on my HMIs as well. Also I animate the valve solenoids and spools and by clicking on the valve a manual popup will appear for operations where it is safe to manually over-ride a valve from the HMI. We have actually talked about modifying the popup window so that not only does it give you manual controls, it gives you a photograph of the actual valve on the machine and part number information so that a maintenance guy can find the physical valve, however no one has the time to do it. My valve ID tags are the same on the drawing as they are in the PLC program and the hydraulic drawing has the IO address of the solenoid valves noted next to the solenoid.

[TimWilborne 108]

I tried putting pictures of the actual components on the Panelview once but it didn't seem to get used very much so I took it off. I am putting part numbers of the wear items such as the hydraulic filters, water filters, and welding consumables that way the wrong part doesn't get put in. This is especially a concern for the fine filters of the hydraulic servo. I am also implementing a stricter PM program for the machine and I plan on doing it to all machines if it works out. All PM will be programed into the PLC. Details of what must be done will be on the Panelview. Whoever does the PM will enter their code to say that it has been done. This information will be emailed to the maintenance supervisor so he can keep track of it. If the PM is not done with in a certain amount of time the machine will signal an immanent fault code which will alarm everyone. If it goes a certain amount past that it will signal a major fault shutting the machine down. It will probably only take one shutdown for whoever is responsible for the PM to realize he better do it or that is our hope. And of course there will also be physical inspections to make sure the guys don't just put their code in to say that they have done it.

[TConnolly 13]

What kind of a valve is that one which you are calling a servo? Definitions of servo vavle -vs- proportional valve have blurred somewhat and some people use the terms interchangably, but normally when I think of servo vavles I think of a two stage valve with a torque motor and a jet pipe flapper which pilots a larger spool valve. Fluid flow exiting the jet pipe is directed toward two ports called receivers which are connected via internal chanels to each end of the valve spool. A flapper between the two receivers is connected to a torque motor, and by applying torque to the flapper a pressure differential is created between the two receivers, and thus between the two ends of the valve spool. This pressure differential moves the spool. An LVDT gives valve spool feedback and completing a servo loop enabling precise position of the valve spool by modulating the torque motor. I've attached a cross section of one of these. A proportional valve on the other hand usually has a direct connection between a proportional coil and the valve spool. Control of proportional valves is often much simpler, however proportional valves do not respond quite as fast as a true servo valve does. However as valve electronics have improved in the last decade proportional valves have replaced servo valves in many applications, including what we do here. The proprotional vavles are less expensive and less prone to problems from slight amounts of contamination than jet pipe servo valves are. I know of at least one manufacturer who is actually attaching a valve spool to a ball screw and is using an electric servo motor to move the spool.

[TConnolly 13]

Want a sucker bet? If your plant is like 99% of other plants then it will only take one shut down for someone to order you to remove the shut down routine from the software. I have one vacuum pump which while it doesn't shut down, it pesters more and more people with emails until the vacuum pump oil gets changed. A shut down hits the bottom line and would mean my head. But pestering emails mean the managers get after whoever is supposed to change the oil.

[TimWilborne 108]

It is your definition of a servo. HERE is the manufacturers literature I think the manifold is what throws you off. We used to have Moog valve on it and we had a lot of trouble with contamination. When going to this valve many didn't want to quote an alternative because of the weird manifold. Which if you are looking at it the schematic looks to be incorrect since the two valves can't function together. Maybe Peter can explain to us why a servo can't do the function of the large valve and provide fine control

[TimWilborne 108]

I don't think it will go that far. This machine is extremely tough and is build with plenty of excess capacity. It could go for a long time without any PM. In fact when it was pulled out of service we figure it was running at about 1/4 of capacity and still had enough power to do the job. Once a machine is in that bad of shape it is much harder to fix. That is why PM is so important. Everyone shouts for PM on machines that break every week. The ones that don't break get ignored.

[Peter Nachtwey 15]

There are big servo valves like the Moog 79-200 and the smaller 79-100 that can both big and fast. I don't think the big valve had to be a very high quality one in this case because it is just jamming the material togehter. Why is there a pressure reducer in this curcuit? The HR Textron valves are not traditional servo valves but the ones with the servo cut spools respond like a servo valve. I never liked the term proportional valve. What is proportional to what? It is the end result that counts. Look at the Bode Plot of the TWControl's HR Textron valve. Notice the response is flat out to about 50 or 60 Hz and the phase lag is only about 40-50 degrees. That is pretty good. FYI, be careful when comparing Bode plots. Notice that this one is rated at a 25% command signal. The response would be much worse at a 90% command signal. You have to know how you are going to use the valve to be able to interpret the Bode plot correctly. You don't want to compare apples and oranges.

[TConnolly 13]

We've got some very big servos on some large pistons (48") and they seem to do just fine. I'll go along with Peter's assessment that finesse is not needed for the DCV and so it appears that this circuit is designed for economy. Contrast that to some of our large servo valves which cost more than several dozen of those size 25 DCVs. (I'm just guessing thats a size 25 DCV from the apparent scale in the picture, I could be off.) ------------------------- BTW Tim, check your PMs. Edited 23 Aug 2006 by Alaric

[TimWilborne 108]

Peter I am going to have to look into that pressure reducing valve some more. I'm not sure why, it could have to do with the machines original controls but here is my first thoughts 1. The original valve was a Moog 74-104. I can't find any specifications on it and can't imagine this being a problem but could it not handle 1000 psi? 2. I think this is a copy of a smaller version the company made. I can't imaging it making a difference but perhaps they needed the larger upset force but wished to keep the servo valve at the original psi 3. I found one of the original setup sheets when I was looking for some drawings. Back then it did not have a servo controller. distances were mechanically adjusted for the different setups. I believe that there was a PSI setting although I can't think of why. Let me do a little digging on that for a few days. Very good question though. I may be overlooking something important. They wouldn't put it there for nothing Good explaination for the reason for the dual valve setup. The cost of that huge servo would be much more than the small servo and big valve. I believe you are right About the Textron servo valve. We replaced the Moog valve several years ago. The Moog valves would only last a few months before trash would clog them up. I do not like the Textron valves performance as much. It does not have the response that the Moog had but it is acceptable. The good thing about it is you could dump saw dust down it and I think it would keep working. It is very tough