VIAIR 250C-IG Compressor

[Al Skierkiewicz]

Originally Posted by ArtemusMaximus

I kinda knew you are going to say that and suspect rules won't change much in regards of pneumatic system.

Be sure to read the 2015 rules before you build next season.

[ArtemusMaximus]

Originally Posted by Al Skierkiewicz

Be sure to read the 2015 rules before you build next season.

That's a long wait

On another note, I am getting discouraged by the prices of aluminum here. To make 3"x3"x0.75" aluminum plate with 1" hole I was quoted $91.04

[cbale2000]

Honestly, I'm still of the opinion there is no need to use a heat sink on the compressor if you're blowing fans over it. Unless you're running the compressor constantly and want to keep it at room temperature, a single 120mm fan (or several smaller ones) blowing across it can keep the compressor reasonably cool under most uses.

[ArtemusMaximus]

Originally Posted by cbale2000

Honestly, I'm still of the opinion there is no need to use a heat sink on the compressor if you're blowing fans over it. Unless you're running the compressor constantly and want to keep it at room temperature, a single 120mm fan (or several smaller ones) blowing across it can keep the compressor reasonably cool under most uses.

Needless or not, my brain is on a roll and I want to finish my design.
Here is idea for the cylinder's heat sink:

Heatsink is made of these

It will require just basic tools to make it.

Also, compressor 3D model I found is not exact measurements and it has 3 bolts not 4 that holding cylinder's heatsink with manifold. So I may have to make model myself.

If I get to the point of making it, I will also do the comparison graphs with temperature readings.

[ArtemusMaximus]

So I did my own the compressor 3D model (Creo 2.0) taking measurement from actual compressor. There still some tweaks left to do but here it is for you assessment:

[Al Skierkiewicz]

Art,
I think if you do an evaluation of the compressor, there will be significantly more heat generated in the head than in the cylinder walls. I am expecting to see more heatsink connected to the head next season.

[ArtemusMaximus]

Originally Posted by Al Skierkiewicz

Art,
I think if you do an evaluation of the compressor, there will be significantly more heat generated in the head than in the cylinder walls. I am expecting to see more heatsink connected to the head next season.

Al,
What in your opinion would be a good test setup? Do you think (compressor--> pressure release valve --> 2 air tanks) would suffice?

In our experience existing heat sink (a.k.a. compressor head) with small fan (40mm zip tied directly on it) was doing decent job cooling it. However, rest of the compressor still heated up a lot. Idea is to have this new heat sink + heat sink for the motor body would + 2 80mm fans dissipate a lot of heat and thus increasing duty cycle. As I stated before, another my objective was to design heat sink in such a way that would require only basic tools to make (hole cutter drill attachment, metal saw, and may be dremmel) as we dot have access to CNC nor to even basic milling machine. Considering that objective I knew that it may cause less efficient heat sink design because it would be harder to make more elaborate shapes of pieces. However, if it will be efficient enough, than ANY other team would be able to produce it.
At the moment best idea I came up with is to use 1/8" x 1.5" x 8' aluminum flat bar easily available here for $20. for this heat sink I need 14 pieces 3" length, that makes 42 inches. lets say 48" with some waste - that only $10 worth of materials.
Motor body heat sink would need a bit more. But it still falls under $40.

[Al Skierkiewicz]

Art,
Not being a mech eng (I play one on TV) I am guessing a lot of what we see is do to the minimal thermal mass in the structure of the compressor. If we add to the thermal mass, then we also add to the weight. The new compressor is coming with a stainless steel, woven hose that is required for operation. I was wondering if a metal tube that had fins on the outside could remove sufficient heat to protect the tubing. We had a rather large shop compressor where I work that had such a pipe. It was a two stage compressor where one piston fed a second piston to get the pressure up. The cooling pipe kept the second piston from heat damage and the flywheel doubled as a fan. The finned tube sat in the draft from the flywheel for forced cooling.

[ArtemusMaximus]

Originally Posted by Al Skierkiewicz

Art,
Not being a mech eng (I play one on TV) I am guessing a lot of what we see is do to the minimal thermal mass in the structure of the compressor. If we add to the thermal mass, then we also add to the weight. The new compressor is coming with a stainless steel, woven hose that is required for operation. I was wondering if a metal tube that had fins on the outside could remove sufficient heat to protect the tubing. We had a rather large shop compressor where I work that had such a pipe. It was a two stage compressor where one piston fed a second piston to get the pressure up. The cooling pipe kept the second piston from heat damage and the flywheel doubled as a fan. The finned tube sat in the draft from the flywheel for forced cooling.

Al,
I am not mech eng either. I have brains for it, but don't have the required education .
VIAIR 250C-IG Compressor is a good option, but at extra 4 lbs weight and extra cost it may not work for all teams.
I haven't done the math, but I am positive that heat sink I am designing would weigh less than 4 lbs.
I haven't learned enough CAD to properly heat stress test , which leaves making it and testing as the only option.

[adciv]

(Computer engineer here) Looking at your design, there's two things that come to mind.
1) Could you design it in two parts that are then screwed together? This would allow you to add it to a compressor without removing the head. (This appears to be required under the design)

2) I think your fins are thicker than they need to be for cooling. Are they there for mechanical support or manufacturability? I'm trying to compare them to higher end computer CPU coolers which use thin copper plate.

[Tristan Lall]

There are several tradeoffs involved in designing heatsink fins, but the main one is providing enough conduction to supply the fin with heat, and enough convection to remove the heat from the fin. (Radiation is often negligible, unless you're in a non-convective environment, like space.)

Convection is affected by fin spacing and airflow. In general, active cooling with a fan can use more closely spaced fins. Passive cooling using natural convection tends to employ widely spaced fins. The most efficient fins in natural convection are often cylindrical or conical, but those are fairly difficult to make, and so straight fins are substituted instead.

Additionally, conduction is affected by the thermal resistance over the material interfaces. If the heat has to cross a lot of material boundaries to get to where it's convected away, the heatsink will tend to be less efficient.

You'll also find that some areas of a heatsink are less efficient than others, relative to the mass of the material used. The squarish blocks on the corner of your design come immediately to mind: they're going to be relatively cold (because they're far from the heat source and exposed to the air), and thus have a small temperature difference with the surroundings, which means less heat flow.

[ArtemusMaximus]

Originally Posted by adciv

(Computer engineer here) Looking at your design, there's two things that come to mind.
1) Could you design it in two parts that are then screwed together? This would allow you to add it to a compressor without removing the head. (This appears to be required under the design)

Yes, that was the plan all along, so there no tinkering with compressor itself. There are 2 parts for Compressor head heatsink and another 2 parts for motor body heatsink.

Originally Posted by adciv

2) I think your fins are thicker than they need to be for cooling. Are they there for mechanical support or manufacturability? I'm trying to compare them to higher end computer CPU coolers which use thin copper plate.

Fins are thicker for manufacturability reasons (material: 0.125" x 1.5" x 8' aluminum flat bar) and also so they can absorb the heat. With CPU heatsink there is always thicker part that absorbs the heat and thin fins to disipate the heat. I haven't yet figured out a way to manufacture such heatsink with basic tools.

[ArtemusMaximus]

Originally Posted by Tristan Lall

There are several tradeoffs involved in designing heatsink fins, but the main one is providing enough conduction to supply the fin with heat, and enough convection to remove the heat from the fin. (Radiation is often negligible, unless you're in a non-convective environment, like space.)

Convection is affected by fin spacing and airflow. In general, active cooling with a fan can use more closely spaced fins. Passive cooling using natural convection tends to employ widely spaced fins. The most efficient fins in natural convection are often cylindrical or conical, but those are fairly difficult to make, and so straight fins are substituted instead.

Additionally, conduction is affected by the thermal resistance over the material interfaces. If the heat has to cross a lot of material boundaries to get to where it's convected away, the heatsink will tend to be less efficient.

You'll also find that some areas of a heatsink are less efficient than others, relative to the mass of the material used. The squarish blocks on the corner of your design come immediately to mind: they're going to be relatively cold (because they're far from the heat source and exposed to the air), and thus have a small temperature difference with the surroundings, which means less heat flow.

Thank you Tristan for your input and good overall summary of the physics involved.
I started working on it over 3 weeks ago and it hasn't been easy process.
I challenged myself with few parameters and as anything else in life few of them competing with each other:

• Cheap. So that even "poor" teams can afford it.
• Easy to manufacture with basic tools. So that even poor teams can afford it. (drill, wood hole cutting attachment, metal saw or grinder with cut wheel etc.)
• Hopefully increase Duty Cycle from 9% to at least 50%
• Easily mounted, so that compressor is untinkered ( to comply with FRC Rules)

Challenge will be making round cuts so they stick tight to the compressor, so that heat can be absorbed more efficiently.

So if any of you have any suggestions to improve the design, I am all ears

[ArtemusMaximus]

Originally Posted by Mark Sheridan

Do note the Viar 90C performance is at 13.8 volts. The Thomas 215ADC38/12 is quoted at 12 volts. Thus the 90C has better performance due to the higher supply voltage in its testing. It would be nice to get data points at 12 volts.

Anyone have a flowmeter handy? Maybe just time the two compressors to fill a known volume would be enough.

http://www.viaircorp.com/90C.html#tabs-2
http://www.cfpwarehouse.com/thomas/c...215adc3824.pdf

Mark,
I tried to look for air flow meters but they are expensive. Our team doesn't have enough funding to justify such a purchase. I wasn't able to find rental locally for the meter either.

I also contacted VIAIR and asked for CFM performance table @ 12V. here is their reply:

... we do it that way since most people use our compressors in vehicles which supply 13.8v from a 12v battery source ...

Hello again,

All of our test machines here strictly run on 13.8v so that everything tests out the same so I can’t provide the CFM at 12 volts.

Sorry about that.

Let me know if you have any other questions.

[Mark Sheridan]

Originally Posted by ArtemusMaximus

Mark,
I tried to look for air flow meters but they are expensive. Our team doesn't have enough funding to justify such a purchase. I wasn't able to find rental locally for the meter either.

I also contacted VIAIR and asked for CFM performance table @ 12V. here is their reply:

Well I hope no team wants to buy ones for such a singular use. I use a mass flow meter (~$800) at work but i am using it for important stuff. I don't want to undo the fixture to test just a compressor.

I bet most teams would just care about how fast a compressor tanks to filll 4 air tanks. I might get around to this if I can buy a thomas 215 this summer. The volumes won't be totally the same, thus the test won't be totally repeatable but I think if the data is that subtle and the compressors are roughly the same, i personally would not care to be more accurate and go with the cheaper compressor.