The hose issue has raised its ugly head again. 1

[PNachtwey]

Does any body know how much hydraulic hose expands as a function of pressure or where I can find such specifications? How can one design without specifications.

I have yet another customer that failed to understand how much hose between the valve and the cylinder affects performance.

If I could get a number in % per 1000 psi or some such number I would be happy.

The bulk modulus of oil is 200,000 psi which basically means the oil will compress 0.5% when the pressure is increased by 1000 psi. If hose expands by 2% for an increase of 1000 psi then I can simply find the equivalent volume of the hose by multiplying the volume by 4. I just need to get in the ball park. Anything close is better than ignoring the problem.

Peter Nachtwey
Delta Computer Systems

http://www.deltamotion.com

 

[MikeHalloran]

This is why hydraulic hose is available with multiple layers of braid or wire spirals.

Surely the effect is much larger than the elastic modulus of the fluid, even for hose that takes three men and a boy to bend. ... but I don't have a quantitative number for it.

Maybe somewhere in the fine print of Gates or Parker documentation...



Mike Halloran
Pembroke Pines, FL, USA

 

[hydtools]

Maybe if you can find the complete paper, a part of which is here:

http://www.md.kth.se/~janh/Hoses/FrequencyResponseofHydraulicHoses1

Ted

 

[kcj]

Peter:

Not sure how valuable this will be but it's all the info I have: In the 1980's, dealing with this same issue, I got some info from Aeroquip on 2781 two wire heavy jacketed roughly 3000 psi hose for open loop cylinder circuits. One was -6 or -8 and the other was -12 or -16. I don’t recall if they had done a past test and dug out the info, or did it at our request. The test was simple, but fairly real world applicable: About 50 feet of hose with assembled fitting at each end was filled with fluid, vacuumed to remove air, then pressurized. The volume of fluid added vs. pressure was measured, and added volume/original volume vs. pressure was determined. This accounted for both area expansion and length contraction of the hose and compression of the fluid inside it (petroleum oil at that time).
I knew enough at the time to know the info I needed, but not enough to ask these questions: how accurately was the hose volume determined, how effectively was the air removed, and how was the added fluid volume determined while still under pressure (I assume a piston area times stroke = volume added.)
When I talked to AQ later, there must have been discussion higher up and this information should not have been given out. Not that it was secretive information, there just were so many variables that AQ did not want someone taking rough test data with an accuracy of +-50% and thinking it was a product specification to be used for accurate design. I was thankful as it was way better than no info at all.

So, given all that disclaimer text, what I got from AQ was a rough bulk modulus of 90,000 to 30,000 psi for smaller and larger hoses respectively. Again, that included the fluid inside it.

Since then, for lack of better info, I have used 100,000-200,000 psi for oil, 100,000 psi for water glycol (because of more of a tendency to dissolve air I think) and 5 to 8 x 10E6 for steel tubing.

I would love to hear if someone has newer or more accurate or even just different information.

kcj

 

[micalbrch]

Of course it depends upon the hose material. But I might have something for you. Look at the attached file, page 122.

 

[PNachtwey]

That is what I need but the units look screwy. cc/ft is a mix and match of SI and British units.

What is interesting is that the expansion is a lot more linear than I would have thought from reading Hydtools document. Hydtools document would make you believe that the pressure change with volume is non-linear and also suffers from hysteresis.

Tonight I will use the inside diameter to calculate the volume for a foot of hose and figure out what percent the cc change is. Fortunately Mathcad can make quick work out of unit conversion. I will also figure out how this data makes sense when simulating and designing.

At least now I know that this data exists from Eaton.

Peter Nachtwey
Delta Computer Systems

http://www.deltamotion.com

 

[kcj]

I can't find the 3130 ID, but if -06 is indeed .375 ID, then it is 1.325 in3 volume per foot of hose = 21.7 cc/ft

Yet expansion at 2000 psi lists at 7 cc/ft. 7/21.7 is about 33% expansion at 2000 psi. That seems really high.

 

[hydtools]

3130 is on page 9.
It is SAE 100R7 hose. Polyurethane cover, synthetic fiber braid, nylon tube.
-06 is .375in. ID

Ted

 

[PNachtwey]

Yet expansion at 2000 psi lists at 7 cc/ft. 7/21.7 is about 33% expansion at 2000 psi. That seems really high.

Why are you shocked?

The hoses in the pdf file are limited to 1/2 in diameter which isn't big enough for many applications. I bet those hoses are just as bad as the 3130 hose. At least you know why I get wound up when people call me and complain that we aren't holding the chipper head positions constant. Using the same fudge factor I used below you would need to multiply all hose volumes by over 32 to get the equivalent volume of oil in a cylinder.

So what does this mean?
Oil compresses at about 0.5% per 1000 psi
I am looking at the 3800-08 or 0.5 inch diameter hose on page 123 because it looks like the stiffest and probably the size most like used between a valve and and a cylinder. The volume per foot is 38.611 cc so at 3000 psi the volume change is 2.5c or 6.5% per 3000 psi or 2.2 % per 1000 psi. Essentially this means that one can find a rough design fudge factor by multiplying the hose volume by roughly the ratio of 2.2%/0.5%=4.32 time the hose volume and add to the other oil volumes.

Jack Johnson used the term hydraulic capacitance which is ?V/?P. Capacitance has units of length^5/force
For for the 3800-08 hose the capacitance per foot is 2.5cc/3000psi=5*10^-5 in^5/lbf or 3.684*10^-14 m^5/N per meter
Mathcad does the unit conversions for me.

Years back someone asked about how much a cylinder expands and I computed that it was so small that it can be ignored because it is much less than the uncertainty of knowing the true value of the bulk modulus of oil itself. The same goes for piping.

I believe it was Ed Danzer that provided some of the info for this because he needs to know how cylinders expand and seals still do their job. I was after the cylinder expansion data for the same reason as here. Thanks to Ed's data I decided not to bother with pipe and cylinder capacitance, I only care about how the oil in them compresses.

Now one can compute how fast the pressure will rise with a little flow of oil.
Ct=Co+Ch // Ct is total capacitance, Co oil cap.. Ch is hose cap...
dP/dt=Q/Ct // Q is flow

I gave micalbrch a star for the info. I am still evaluating Hydtools document.
What about bigger diameter hoses?
What about other manufacturers?

If all of you would ask your suppliers for this information for the hose or any component you use we could build a data base. It would be nice to have a specification for the capacitance per ft of hose or per meter of hose when using SI units.

The document that hydtools still bothers me in that it indicated the problem is much more complicated than what the Eaton document says it is. I also doubt the hose expands as linearly as what is said.


Peter Nachtwey
Delta Computer Systems

http://www.deltamotion.com

 

[kcj]

Yet expansion at 2000 psi lists at 7 cc/ft. 7/21.7 is about 33% expansion at 2000 psi. That seems really high.

Why are you shocked?



not shocked that hose hurts, and I agree with all you wrote. (We have to control a high momentum load with a total of about 20 ft of -06 hose between valve and cylinder and it is not fun. Have had to do it for many years because of environment, and it works ok, but it is NOT good official design.)

What I mean was that the chart on pg 122 shows (using -8 now instead of -6 in my quotes) about 10 cc/ft at 2000 psi. 10cc/ ft / 38.6 cc/ft volume is 26% volume expansion. That is what I questioned and think is unrealistically high.
2000 psi/ .26 is bulk mod of 7600 psi, seems way low. (Converting to capacitance is just units. I am just more able to visualize in terms of 7600 psi))


You quoted 2.5 cc./ft at 3000 psi. Where is that from-I can't see the -08 plotted above 2000 psi. the -03 at 3000 is about 2.5 cc. Am I reading the chart on pg 1223 wrong? appears to be in units of cc/ft and psi, but.....

However, using the 2.5 cc at 3000, 3000/.065 = bulk mod of 46,000 psi. THAT would be what I would expect for a -08, maybe even better than I'd expect for a synthetic hose.

So, my comments meant that we all agree hose is way bad, but I don't think the Eaton data seems right. If you got data of 2.5 cc at 3000 psi, that seems reasonable. 4.32x tubing sounds right, but 32x tubing seems off.

 

[PNachtwey]

kcj, I am looking at a different hose than you. I am looking at the 3800-08 in the middle right of page 123. It is much stiffer than the hose you are looking at.

I agree with your value but I think you need to think in terms of capacitance.
The bulk modulus of the oil inside the hose is still going to be that of the oil which is ideally about 200,000 psi. The capacitance of the oil is the bulk modulus of the oil/volume. The bulk modulus of the hose is the change in volume over the change in pressure. What makes this a little difficult is that the volume of the hose is always changing so the capacitance per length of hose for oil is changing as a function of pressure. The two capacitances for oil and hose must be added together.

Doing it your way it it would be simpler to not that oil compresses 1% per 2000 psi and the hose expands 26% per 2000 pounds so the volume of the hose needs to be 'normalized' by multiplying it by (1%+26%)/1% or 27 times and then add that volume to the volume in the cylinder or hard pipes and use the just the bulk modulus of oil. Either way will work.

Note that the expansion of the hose is per foot!!!!

"but I don't think the Eaton data seems right. "
I would like to see the data from other manufacturers. My big pet peeve is that the manufacturers don't supply adequate information.

I would also like to know the pressure drop per foot as a function of flow. This would give me a 'resistance value'.

Peter Nachtwey
Delta Computer Systems

http://www.deltamotion.com

 

[Gkranz]

Peter,
I see you are at it again regarding hose expansion and I recall the many discussions we have had over the years. I normally do not post on these forums, but do like to read them, and this one I just could not pass up, ha, ha.

Look guys, most of the data suggested is suspect at best, and is hose specific. There is good reasons for this, which include size, temperature, rubber compound, length, all are variables you cannot control Peter. Just because your controller can measure out 10 decibel points, does not mean the real word behaves the same as your simulator does.

As you know I was doing servo systems years before they became the latest fashion, and we always told the customer the best location of the servo valve was attached to the actuator. However, sometimes in the real world this is not possible, so when we did need to locate the servo away from the actuator we use tubing or in some cases hose.

Yes, that’s right hose, however we did two things and that was to increase the oil velocity to around 35-40 feet per second, and use 4 to 6 spiral hose because it expanded the least, if at all. The expansion rate was so small it was not worth calculating and had little to no affect on the overall performance of the system.

Now, I’ll give you something to really work on old buddy. Oil expansion and oil compression, both are bigger factors that will affect performance. Also your Bulk Modulus being stable at 200,000 psi is wrong in your calculations. It will change with temperature and pressure, which you need to account for in your program. I have that information if you want it, just let me know or maybe I will post it to my web site (westerndynamics.com) and you can get it there. Always good to read your posts, but I need to keep you connected to the real world and away from that computer, ha, ha.


Best Regards,
Westerndynamics.com

 

[PNachtwey]

Thank you for your input.

Peter Nachtwey
Delta Computer Systems

http://www.deltamotion.com