Story of Slow Tomato Belts by Insane Hydraulics

     Two years ago, in the Summer, I got an "S.O.S." call from a client, who provided field assistance to agricultural equipment in his area. For several days he'd been "struggling" with a big tomato harvester and was starting to become desperate (both he and the client of his), as the downtime was very expensive and the problem remained no matter what he did. It appeared that he new these machines from head to toes and had successfully provided assistance to many clients in the region for years.

     The machine owner was complaining that the tomato transporting belts were running slow, which itself seemed a simple problem and even the complete documentation with all hydraulic schematics included was available.

     For those of you who are not familiar with this type of equipment a short explanation: these harvesters are quite complex machines, with lots of hydraulics and electronics, and the pneumatic part that separates green tomatoes from red is just spectacular. Normally these machines are equipped with several transporting belts directly driven by gerotor type hydraulic motors connected in series. This one had seven Danfoss OMR motors driving seven belts, six of which were equipped with speed regulating valve (schematics). The series connection allows to spare space, hoses and pump size, as these belts are mostly horizontal and don't require producing much torque (it is the last motor in the series that produces the most torque, as normally it drives the vertical belt carrying tomatoes to the tomato-transporting truck). Very often manufacturers won't even install drain lines on the motors, which was not the case of this machine, though.

     Although I am yet to find a Portuguese farmer who is completely satisfied with the speed of his machinery (faster, damn you, you stupid machine, faster!!!), this mashine indeed was running slower than it should. The mechanic had applied the good old "let's replace the old stuff with new and see what happens" principle, and apparently it didn't work, as practically all of the circuit had been replaced, the pump had been opened AND tested in our workshop and did not represent any type of excessive wear, and them damned belts remained slow.

    Simple things first - pump actuation, is it belt driven? No, it is a directly connected gear pump with a spline shaft. O.K., moving on, although the client had verified the relief and the by-pass valves, I verified them again (don't trust anybody, remember?), in fact I made a direct connection from the pump to the first motor as it was the simplest and the easiest manner at the moment due to the hose and the fitting size. Nothing changed. (Children, don't try this at home!) Then I disconnected the drain line of the first motor to check for excessive leakage, the oil was trickling from the drain line, as it should as the motor was almost new. (By the way the mechanic had verified the drain lines of all the motors and did not detect anything unusual) At this point I was sure that at least the first motor had a good volumetric efficiency, and as it was connected directly to the pump I decided to use it as a flowmeter, (the mechanic had a digital flowmeter at the site, but he claimed it was broken, cause when he had tried to get a reading of the flow at the pump's outlet, the flowmeter showed very high flow rate deviations, sometimes two or three times as big as the pump's nominal flow).

   A small strip of reflecting duck tape on the motor's shaft and an optic pick-up rpm gauge and voila - a simple and effective flow meter without any oil on my clothes (wish it always were like this...). The machine started and then I detected something that was definitely not normal. As the diesel RPM was climbing, so was the flow, until a certain point, and then the flow increase (hydraulic motor RPM) would seem to stop being linear with the diesel RPM, in fact it seemed to remain at almost the same level. Meaning that when the diesel went from 1000 RPM to 2000, the flow was to double, and this was NOT happening. Why?! A quick verification of the diesel flywheel speed confirmed that the machine's tachometer reading was accurate, and then the machines owner added that the belts were slower in the morning and when the oil was at normal temperature the speed improved a little.

    STOP! At this point I was pretty sure where the problem was. If you don't have any clue, read the above and think a little, you already have all the information to indicate the most probable cause of the problem. Any ideas?

    O.K. Let us first state the obvious:

1. The first motor has good efficiency, hence the motor's speed is directly connected to the oil flow passing through it. When the motor's speed stops increasing it can only mean one thing - the oil flow stopped increasing. So the main reason for the belts being slow is insufficient oil flow.
2. The pump is a fixed displacement type and was proven to have a very good efficiency, the mechanical part which moves the pump is o.k., and the pump's outlet is connected directly to the first motor of the series with no oil deviation whatsoever.

     In this case, the ONLY way to make the insufficient oil flow to come out of the pump is not to let the sufficient amount of oil to ENTER the pump. Which means the deficient suction line (partially obstructed or collapsed) and explains all of the above symptoms - belts running slow, oil flow stopping to increase at certain point (when the pressure in the suction life drops to evaporation point at the given temperature and the pump instead of liquid oil starts pumping oil/vapor mixture, causing noise and cavitation at the same time), explains why the speed improved with hotter, and therefore more fluid oil. It even explained why the mechanics couldn't get an accurate reading from his digital flow-meter which he had installed at the pump's outlet. This phenomenon happens with turbine type flow meters (most common) when you measure flow of a gas/oil mixture. The turbine wheel is very light, and the presence of gas/vapor pockets in the flow causes its jerky movement, which the electronic counter interprets incorrectly. On the other hand, the heavy and belt loaded gerotor type motor's rotation speed will generally correspond to the amount of oil passing through it, with the vapor pockets imploding upon entering the motor.

     Another possible explanation I thought of at the moment was air entering in the suction line (which would also mean deficient suction line), but it had to be a pretty fair amount of air, and this possibility was quickly discarded when I examined the oil coming out of the motor - it was clean. Were it to have a significant amount of air bubbles in it, it would come out as a nice beer foam colored mixture (ah-gh-gh-gh-gh-gh, b-e-e-e-er (to be pronounced with Homer Simpson's voice)). So I was left with only one theory - strangulated suction line, which , by the way, due to the machine's design, was a 3 meter long R1, 1 1/4 inch size hydraulic hose.

     However when I put my idea of what was happening in front of the client he was rather sarcastic about it, and found that it was practically impossible for the suction line to be obstructed. He claimed that thorough tank cleaning had been performed and the suction line hoses were almost new.

     To prove my point I would need a vacuum gage installed in the suction line, but I didn't have one at the site, and the mechanic (to my surprise) did not know what a vacuum gage was. A quick look around the warehouse (in the middle of nowhere) revealed an old air compressor, with a 5 bar pressure gauge. I got the permission from the owner and started "severe gauge violation", wish I had my camera on me at the time! What I did was pretty simple, actually, I opened the gauge and carefully moved the arrow, so that its resting point would be somewhere in the middle of the scale, thus allowing it to read negative pressure (in relation to atmospheric, I know, yes..). Then I tested it with my mouth, don't laugh now, your mouth can actually create around half a bar negative, so it is a good way to test vacuum gauges in case of doubt, have you ever let a beer bottle get stuck on your tongue? Try the same with a vacuum gauge inlet and check its reading! Don't let anyone see you do this!

     Then I introduced the client to the "suction line theory", and showed him how the suction gauge worked. Luckily I had all the necessary fittings in my test tools box to connect the gauge to the pump's inlet. The machine started, the arrow jumped down right away, and when the rpms climbed to 1500, the gauge was showing around point seven bar negative, which was wa-a-ay too low! Although the proof was showing the middle finger right to his face, the man was still a little bit reluctant about my theory, so I offered him to test a similar suction line mounted next to this one. Another hour of downtime (HE wanted it!), and the improvised vacuum gauge is connected to the "neighbour" suction line. As I was predicting, when the machine started, the pointer hardly moved from its place no matter what the engine speed was, (like every decent suction line should be). Hurray!!!

     As it was getting late, no further work was done to the machine. I advised the client to inspect the suction line and the tank to see where the strangulation point was. Unfortunately, I newer knew what exactly had strangulated the line, I think they solved the problem by replacing the whole hose. A week later the mechanic bought from us two vacuum gauges and a set of accessories, by the way...

    Oh yeah, one more thing. Some time afterwards the owner of the same machine wanted to increase the speed of his belts even more, around 10 per cent beyond factory values (I truly don't know what is the deal with Portuguese farmers and their machinery...) Some calculations were made, and the conclusion was that the flow, given by the pump, was more then enough. So after inspecting the priority valves (schematics), I found that the needle valve, even when totally opened, would still strangulate the oil flow enough to move the flow dividing spool, due to the rather small diameter of the seat (around 4.5mm in this case). The simple solution was to drill the seats to a 5.5mm diameter, which solved the problem instantly. (not solved the problem, as there had been no problem with the valves, just gave the client what he wanted).