Coolant Hose Material Selection for 10 yrs No Service 4

[Mark172]

Hi all, I'm a mechanical engineer at a research institution building an array of spectrographs. I'm currently designing the coolant assembly and attempting to specify the coolant lines for a minimum of 10 yrs operation with no line maintenance. The other constraints and requirements are listed below. I'm mid career but have very little experience with long-life high-reliability fluid systems, so I am wondering if any of you can suggest a material for the coolant hoses. These CANNOT leak. I've been looking at EPDM, PVC, PTFE, and copper. I like the reliability and low cost of copper lines but we need some flexibility, as these lines attach directly to a detector package, the location of which is very precisely controlled. I cannot induce any significant force on that connection to the detector (less than 1 lbf).

I spent some time on the phone with a Parker-Hannifin applications engineer in their hosing department, but it wasn't very helpful - no confidence or guarantee for any of their lines beyond 1 year. So any tips on hose material selection you all have would be much appreciated!! Thank you all.


Requirements:
Leak free for 10 years.
No maintenance.
Bend radius of < 4 inches


Environment:
Dry air (desert)
10C coolant temperature
20/80 Glycol/Water coolant
Low flow rate (1 gallon/minute)
Low pressure (<50 psi)
0-30 C ambient temperature
Lines are static, do not move (except for ~50 micron travel at mate to detector)

Blue lines indicated needed coolant lines:

 

[Pud]

I would suggest nylon 12 if you can find it. It's got outstanding chemical resistance.
Car radiator header tanks are made from hydrolysis resistant nylon 6.6.
Alternatively copper with a beryllium copper bellows brazed in a suitable position for the vibration.


Politicians like to panic, they need activity. It is their substitute for achievement.

 

[MintJulep]

Hoses that CANNOT leak are impossible.

Having said that, your application doesn't really seem very challenging. Any number of solutions that "probably won't leak" should be available.

Time for an actual risk assessment.

If none of your constraints can change, then perhaps a co-axial hose with the outer passage connected to a scavenge and detection system would work.

If you are really not as constrained as you state, then perhaps you could use a cooling fluid that is benign to your other equipment when it does leak.

 

[rb1957]

MJ beat me to it ! It is easier to say "cannot leak" than to give some probability for leakage. Like MJ, what happens if there is a leak ? how to mitigate the consequences of a leak ? how to minimise the chance of a leak ?

another day in paradise, or is paradise one day closer ?

 

[TugboatEng]

EPDM is used by all automotive manufacturers and can easily last 10-30 years. Any exposure to hydrocarbon oils will lead to failure.

Silicone would provide an even longer life and is readily available in blue. Silicone hoses affected by oils. Silicone hoses are very permeable.

Both materials have good UV resistance.

We use hose from Federal Hose for our engine applications. We have many that have been in service for 10 years and we see no deterioration.

https://www.federalhose.com/

 

[Mark172]

Happy to get into the details - 10 years, no leaks, no service are the requirements handed to me for the coolant assembly. Yes that's manager speak, in terms of probability at least 3 sigma odds of no leak across 10 years and 24 cameras (48 total lines for the cameras plus 16 more for cooling the power supplies).

We are very averse to leaks. There are sensitive ($) optics, custom detectors ($), and high voltage in close proximity. The instrument itself will be located in a remote location (Atacama desert of Chile) and the logistics of repair are challenging, especially in the COVID age. Many aspects of the array design are for leak contingency, such as all lines being on the nadir side so that any leaks will fall away from the optics and the majority of the electrical connections at the top. We have flow sensors on each line, humidity sensors, as well as temperature sensors that are all monitored in software which automatically cuts flow if they sway out of paramater range. We have watched other instruments suffer glycol leaks with one being especially nasty resulting in a very costly repair and downtime on the order of a year.

As MJ said, the 10 yr no leak requirement doesn't seem too challenging given the low pressure and quite benign environment. A coaxial hose sounds nice but well out of our budget. I'll look into Nylon 12. What are your thoughts on Parker 801, a buna-n rubber line with a polyethylene coating?

 

[rb1957]

ah ! so now you want impossible performance and within some arbitrary budget (smile, that's business as usual).

Roll the Billy Connolly Project Manager sketch (on YT).

another day in paradise, or is paradise one day closer ?

 

[TugboatEng]

OP is willing to pay the premium for Parker hoses so they aren't going full cheap.

I would avoid Buna-N. It's life starts to get very short once temps climb past 150F. It's even shorter in water.

What diameter hose are we talking here?

 

[Mark172]

Ah rb, I wish it stopped at impossible performance and budget demands. Did I mention they want it yesterday? Oh well, this is our lot.

Thank you tugboat for the perspective on buna-N and the EPDM recommendation. I'll look into EPDM options. This is for a 1/4" ID.

 

[EdStainless]

Is there a reason that you are not just running them in metal?
1/4" SS tube isn't very expensive and fairly easy to work with.
Buy it from you Swagelok distributor.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

 

[Mark172]

SS and copper are in the design space, but we need a flexible connection to the detector. That would require a bellows of some kind, which introduces a break in the line and additional connectors that work against the requirement to minimize leak risk. Furthermore such bellows are ~$100/each which gets pricey quickly.

Frankly I would prefer copper lines, brazed to a bellows as pud mentioned above, but it may be overkill and is almost certainly too expensive. I'm leaning toward Parker EPDM hoses at this point...

 

[MintJulep]

As Mark172 explained in the original post, they need to limit the force transmitted from the coolant lines to the equipment.

Hard tube with a flexible hose at the connection could do that, but the fitting or connection would introduce a potential leak point, and the application is not tolerant of leaks.

I might be inclined towards hydraulic hose. Because it comes with good options for fittings.

 

[TugboatEng]

If your water is treated for corrosion resistance there isn't a hose system that won't make 10 years in your application. There are certainly some steps you can take to ensure this to be the case. EPDM is really the best choice for long lasting hose provided there is no exposure to oil.

Avoid repurposing hoses. Hydraulic hoses are strong and all but the fittings may not have the corrosion resistance you need and the 37 degree flare connections are prone to leakage.

You can make metal tubing flexible by bending a coil in to it. No need for pre-fabricated expansion joints. Do not try to flare stainless tubing, use compression fittings such as Swagelok.

 

[EdStainless]

I have seen SS used in robotic application with large motions by using either coils and "Omega" bends.
And 1/4" SS sure isn't very rigid.
Cu would even be more flexible, even using heavy walls (low modulus).
Personally I would run hard lines to the back edge of the platform and then braze a fitting and run hose to the final connections. This would keep the top area clean and solid, while providing flexibility.
I wholly agree with the EPDM option. Good water resistance, and good UV/weathering resistance.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

 

[MacGyverS2000]

Minimal stress against the equipment might still be achieved with solid lines... you just need appropriate staking of the lines near the equipment entry, typically a solid mount that allows for a bit of adjustment. Surely you can run the entry/exit points down to a hard mounting surface outside each unit (as part of the design) where such a staked connection could be more readily made.

Dan - Owner

http://www.Hi-TecDesigns.com

 

[btrueblood]

Epdm hose is fine...for vibration and small displacements...but if you chafe it (rub against other things) it will fail (like any other hose). If you flex it significantly, the typical hose barb and band clamp or crimped fittings will also chafe and slip, and cause leaks. Ideally you want a hose whose inner liner is bonded during the rubber curing process to the metal fitting...good luck finding that. We went with coiled stainless (hand formed on a wooden mandrel) and swagelok fittings for a system that underwent continuous flexure, it has never leaked, in multiple installations. It does take up more space than a hose, though.

 

[TugboatEng]

If you do go with rubber hose and use hose clamps be sure to avoid the ubiquitous worm drive clamp. They aren't round and don't compensate for compression set as the hose ages. Spring type hose clamps such as you'll see under the hood or your modern car are superior. Pinch type clamp work well for small diameters. Make sure to use the smooth bore style.

 

[Mark172]

Thanks tugboat for the tip on the spring clamps, I have to admit that was a blind spot.

Thank you all for your responses, there is a wealth of useful information here. I should really come here more often.