20 gpm live hydraulics

I'm looking at putting together the last pieces  in adding a crank driven 20 gpm pump to my D-19. I'm looking at keeping the reservoir size to a minimum by adding a cooler in front of the rad. I'd like to incorporate part of the loader frame as the reservoir. Is there a general rule of thumb as to how much the reservoir should hold if I return the oil through the cooler before dumping to the tank?

No It's more for a hoe attachments and small scraper.

Reservoir volume serves more functions than just providing offset for cylinders and cooling time for flow.

Aereation occurs in the fluid as a happenenstance of many things... everything from leaks in the suction path to thermal expansion of the pump parts, valve seals, cylinder seals, etc...  and the ONLY thing that removes air bubbles from oil, is TIME.  This is the first reason why one gallon of capacity for every gallon of fluid flow is the general theme for hydraulic design.  The other things to keep in mind, are thermal expansion (you'll need room for that oil to expand), as well as keeping the reservoir clean and dry.

The latter point is something many don't understand.  Let's say you have a 20gpm system, with a 21 gallon reservoir, and 20 gallons of fluid... there's one gallon of volume available for expansion volume.  Let's say the oil's volume increases with thermal expansion to the value of about a half gallon.  A little aereation brings the level up another gallon of volume, and your reservoir is now burping fluid.

Better way, is to use a larger reservoir... not only to prevent belching oil froth out, but also so that there's a fair chunk of volume IN the tank to allow for the expansion.

Let's say you cap the reservoir tight... no breather...   when the fluid expands, you'll have positive-pressure inside the reservoir, and eventually you'll have valve and cylinder seals  leaking from the residual pressure... but you'll have a reservoir that is otherwise sealed airtight, hence not drawing in contaminants and condensation.

If you INCREASE the size of the reservoir... say... 30 gallons... the amount of pressure change inside the reservoir will be much lower, so the likelyhood of weeping leaks is much lower.  Furthermore, if the tank is equipped with a breather that will relieve at say... a 2psi differential, it'll keep a nice, clean oil, and not weep.

Another thing to keep in mind... aereation is a brick-wall for hydraulic efficiency.  When you run fluid through the pump, it pressurizes to do work.  When there's aereation... even small gas bubbles... entrained in the fluid, the pump compresses those bubbles, and per Combined Gas Law (P=V/T), the TEMPERATURE of the gaseous volume of those bubbles skyrockets.  Let's say the entrained bubbles and fluid are at 100F, and you compress them (both) to 2000psi...  the VOLUME of those bubbles will immediately compress to 1/2000th of their size, and temperature of those air-bubbles will rise by a corresponding figure.  Herr Diesel pays us a visit when this happens...  the oil will actually ignite and 'coke up'.

Most of the time, when someone asks me to troubleshoot a system plagued with high oil temperatures or apparently insufficient cooling capacity, it's usually an aeration or cavitation condition.  Cure that, and the fluid cools down immediately.

So there's never a case in fluid power design where running a small reservoir in a large system pays off.

Now, one thing that HELPS improve system performance in a small-res system, is to utilize a TALL/SKINNY reservoir, with the pickup low, the return about halfway up... that way, air bubbles will be more likely to find their way UP than down... and there's be less likelyhood that returning oil will pull entrainment from above, down to the suction point.

As far as cooling bypass, I personally wouldn't use a thermostatic bypass.  Instead, I would mount the cooler in a position where it's effect would be marginal, until such time that a cooling fan draws cooling air over the heat exchanger.

Another thing to keep in mind... at least for cold-weather installations, that getting the fluid UP TO TEMPERATURE, is every bit as important as keeping it from getting too hot.  Cold startup operations on hydraulic pumps is seemingly nothing short of traumatic.  I design my systems so that when I'm dealing with -30F, the fluid is brought to temperature quickly... at least, on the suction side of the pump, so that the pump's not thrashing away, with oil slowly 'oozing' it's way down a suction tube.  The pump is lubricated by fluid, but during that startup period, the tolerances in the pump are all out-of-whack (thermal expansion works both ways), and the oil's not there to do the lubrication job.

So if you're planning on a system that'll be experiencing high duty cycle and high power levels, I'd recommend doing absolutely everything you can, to assure ample volume not only for your 20-gallons of fluid, but also for expansion volume AND a path that tends to trap bubbles so they don't get sucked back through the pump.

PV=T not P=V/T

First of all, You NEVER want a sealed reservouir, it has to have a vent, and it should be filtered. I you seal it, all your seals will fail.

Keeping mind that the oil is simply flowing from the pump back to the reservior until it's asked to do some kind of work. If you wish to split the flow into seperate valves and they stay split until they reach the reservior then pick which circut will do the most work and route the cooler loop through this side. The oil is placed under pressure when it's asked to perform the work against what ever the applied resistance is. Since it takes lots more HP to work the oil against the resistance you put more HP into the oil. The more HP you put into the oil the more that must be removed. The in-effeciancies of hydraulics are that you can never remove the same amount of HP that was put into it so the un-used HP converts to heat. So, if I explained it so you all could follow me, the more work a circut performs the more heat is created, put that circut through the cooler.

Thanks for correcting the formula, Stan-  I shouldn't type while eating dinner... ketchup on the keys...