"Radius" by BigBen2k

[gone_fishin]

Originally posted by unregistered
I'm guessing that if myv65 can mention Roark, anything goes !
(an ind std ref BTW, a good read but tough sledding for the untutored)

after designing and doing failure analysis of o-ring joints for many years (in both static and cyclic pressure service),
there are several useful guidlines:
- when the component surfaces make contact, as they must, there should be 30 to 40% compression of the crossection
- the o-ring's volume should be between 60 and 70% of the groove's
- using a 30 to 50 Shore A durometer (hardness) material will make the surface finish less critical
- the o-ring's (centerline) length should match the groove's (do NOT stretch)
- the groove must have an outer containment ring to 'capture' the o-ring, else it is functioning as a gasket (different design basis)

do not use a sealant with the o-ring (no need if properly designed, will overfill the groove, make disassembly a pain, etc)

g_f
o-rings, properly designed, are FAR more reliable than gaskets

Hmm.. I won't argue with you there but the task at hand in Cathar's block is critical. The design is based on forced water through the slot and a potential loss of presure from user ignorance is a likely scenario.

[BillA]

relax,
give natural selection a chance

[bigben2k]

(Aaaargh! Lost my post!)

Thanks for the history tip, myv65. Thank you BillA, for the o-ring specs, that was very helpful!!!

I found the Socket A specs .

It is 52.4 mm wide, by 65.5 mm long. As far as I can tell, the clearance between the surface of the mobo, and the top of the socketed area, is 5.50 mm.

It is longer than wider because of what is called the "cam box", which is where the lever mechanism sits. There is a prescribed notch to heat sinks (those that lay over the cam box) in the order of 2.0 mm.

Prescribed clamping pressure, which should be in fact longitudinally applied (so the Paul Vodrazka mount is out of spec) over the core, is between 12 and 24 lbs f, typical 14 to 18, nominal at 16.

I'll try to find a reference to the relative bolt hole position.

[bigben2k]

Originally posted by unregistered
...

After designing and doing failure analysis of o-ring joints for many years (in both static and cyclic pressure service),
there are several useful guidlines:
- when the component surfaces make contact, as they must, there should be 30 to 40% compression of the crossection
- the o-ring's volume should be between 60 and 70% of the groove's
- using a 30 to 50 Shore A durometer (hardness) material will make the surface finish less critical
- the o-ring's (centerline) length should match the groove's (do NOT stretch)
- the groove must have an outer containment ring to 'capture' the o-ring, else it is functioning as a gasket (different design basis)

do not use a sealant with the o-ring (no need if properly designed, will overfill the groove, make disassembly a pain, etc)

Utabintarbo had mentionned o-rings from McMaster ( www.msmaster.com ). I looked through their selection, and it seems that all they have, are o-rings with a shore-A durometer rating of 65 to 90 (55 in shore D, Teflon). We were shooting for a 1/16 round cross-section.

The material selection that they have includes: Polyurethane, Neoprene, Buna-N, Viton, Teflon, Kalrez, Ethylene propylene (EPDM) and silicone.

There's also a slightly different shaped o-ring, but still not within Shore A 30 to 50.

Until I find a supplier, here's a cutsy anecdotal story:
Finding a Shore A durometer

[BillA]

yellow pages Ben, more o-ring suppliers in Houston than you could call in a week

you can have o-rings any way you spec 'em
molded are better, but 'limit' you to stock sizes (which you can design for)
molds are not expensive, but probably not warranted
spliced joint in 'any' size, just do a bend/peel test on the joint
why be fancy ? use Buna N
ck the chem res charts - no biggie

harder take higher pressure, you don't need that
a fatter section gives you more slop in the tolerances -> 3/16"

[bigben2k]

According to http://www.oringswest.com/

Methanol will work with EPDM, Neoprene (Chloroprene), Fluorosilicone and silicone, but NOT nitrile (buna-N) nor Viton.

Water will work with Viton and Neoprene, but only "fairly", which they mean as "Static usually OK". Otherwise will work with all above materials.

Which leaves:
EPDM
Neoprene (Chloroprene)
Fluorosilicone
Silicone

and

Polyurethane
Teflon
Kalrez (tm)


Now let's take out Polyurethane, for it's poor acid resistance.

[BillA]

EDPM

who is actually running methanol ?
not talking, doing eh ?

[bigben2k]

OK, EDPM, Durometer rating Shore-A of 40,

3/16 diameter, ID: 1 3/8, OD: 1 3/4 (Dang, too small!)
AS568A-324

3/16 diameter, ID: 1 1/2, OD: 1 7/8
AS568A-325

Waiting for quote...

(Methanol: with water, Windshield Wiper fluid)

[bigben2k]

Les was kind enough to give me a more accurate overview of the thermal calculations. I'll give it another shot.

In the mean time, here's something to ponder: Copper 110?

[Yo-DUH_87]

Originally posted by bigben2k
(Methanol: with water, Windshield Wiper fluid)

You're going with the windshield wiper fluid afterall?

[utabintarbo]

Originally posted by bigben2k
Les was kind enough to give me a more accurate overview of the thermal calculations. I'll give it another shot.

In the mean time, here's something to ponder: Copper 110?

According to the spec sheet on McMasterCarr's web site, copper 101 (low o2) and copper 110 have exactly the same thermal conductivity with 110 being approx. 50% of the cost.

Ponder that!

Bob

[MadDogMe]

The pads around the core are soft foamy, rubber. they compress alot. I think they're to make sure the HS 'starts off' at a level rather than remaining so...

[gone_fishin]

Originally posted by utabintarbo
According to the spec sheet on McMasterCarr's web site, copper 101 (low o2) and copper 110 have exactly the same thermal conductivity with 110 being approx. 50% of the cost.

Ponder that!

Bob

101 is 99.9% pure and 110 is 99.5% pure. Getting it to 99.9 involve more refinement, hence the higher price. Conductivity is listed as high. There are some alloys of copper that are strong and may be more suitable for an ultra thin base but the tradeoff is slightly less thermal conductivity. The crystaline structure is also affected by the way it was worked to stock form.

Here is an excellent link for all copper alloys, their uses, properties, and forms. linky linky

[bigben2k]

Ok, here's a recap:

We have a final fin pattern, a base material (C110), a socket mounting solution, a top design, a cube-res design (Utabintarbo is drafting it)...

I just gave Utabintarbo the exterior dimensions of the block, which I think should be 52.4 mm by 52.4mm, which is as wide as the socket. This leaves out the socket tab mounts.

Mounting this thing with springs, it seems that the mounting holes have changed from 0.230 inch to 0.150 inch. The spacing of the holes is clear, but I'm looking for the relative position to the socket. AMD specs no longer include mounting holes. clamping force is as specified earlier, so 1/4 max load per spring.


Yo_duh87: Windshield wiper fluid is for phase 3. Note that methanol is hazardous, and shouldn't be used unless one knows exactly what precautions need to be taken.

Utabintarbo, go for C110 (actually C11000).

MadDogMe: I realize that they're soft. Re-reading AMD specs, the pads are 1.9 mm thick, and the core is somewhere around 0.8 to 0.88, so if the core is going to live, they have to squish down. AMD specs also call for a mount that applies a longitudinal load across the core, with no less than +/- 1.5 mm in deviation. I think we're OK, but I'll have to check.

Gone_fishin: nice link!

[bigben2k]

Well, I kindly borrowed this graph from Cathar's block thread on OCAU. For some reason, the dimensions are in inches (?!?).

The weird thing is that the holes are actually off-center. (?!?) Spec calls for no more than 1.5 mm off center, and the holes are 1.25mm off. What where they thinking?

Does anyone have the mounting hole specs? (I mean, do you have an old copy of the AMD PDF?)

Does anyone have an opinion on a socket (tab type) mount?

[bigben2k]

Discussing with Utabintarbo.

The 3/16 o-ring is too big. The bolts wouldn't fit. In order to use a 3/16 o-ring, we'd have to make the block longer and wider, which means stepping over the cam box, which means putting a 2mm clearance notch (wich happens to be the baseplate thickness).

So back to plan A: a 1/16 o-ring. I still prefer EDPM, and I have to point out that buna-N won't work with meth, so it's out, at least for me (but I suppose that I could swap it later).

Got the quote back: "$100". I'm thinking of going with the McMaster offering, with a durometer rating Shore-A of 70.

What I don't get is this:

Originally posted by unregistered
- when the component surfaces make contact, as they must, there should be 30 to 40% compression of the crossection
- the o-ring's volume should be between 60 and 70% of the groove's

If the compression is 30 percent, and the o-ring volume is 70% of the groove, isn't the o-ring filling the channel?

If the compression is 40 percent, and the o-ring volume is 60% of the groove, isn't the o-ring filling the channel?

What's the gap supposed to be, if any?

In the mean time, here's Utabintarbo's latest render (without the top).

[bigben2k]

Originally posted by unregistered
who is actually running methanol ?
not talking, doing eh ?

Meltman is using it, and quite succesfully.
Check it out

[bigben2k]

Speaking of top...

The latest direction is towards copper, for structural integrity.

Previously mentionned was polycarbonate, and acrylic. What hasn't been mentioned yet, is glass. (flameproof suit on! gone_fishin is going to like this!)

Come to think of it Plexiglass hasn't been mentionned either.

[gone_fishin]

Originally posted by bigben2k
Speaking of top...

The latest direction is towards copper, for structural integrity.

Previously mentionned was polycarbonate, and acrylic. What hasn't been mentioned yet, is glass. (flameproof suit on! gone_fishin is going to like this!)

Come to think of it Plexiglass hasn't been mentionned either.

Nooooooooo........... 1/2" poly will not flex at that small of a radius
Just messin, do as you see fit. Kind of overkill on the assembly screws though dontcha think?

[bigben2k]

Better safe than sorry, plus I like the 8 bolt design...

Did your bulletproof glass work OK?

[gone_fishin]

Originally posted by bigben2k
Better safe than sorry, plus I like the 8 bolt design...

Did your bulletproof glass work OK?

Still working as good as the first day they were assembled

[bigben2k]

kewl! Now, tell me again how you got those barbs in the glass?

[gone_fishin]

Originally posted by bigben2k
kewl! Now, tell me again how you got those barbs in the glass?

Instigation

[nicozeg]

Hey Bigben you should reconsider the new cuts on the center fins, it seriously compromise structural stiffness of your block. If you keep connected the central crossing fins they make the base far more "unbendable", not needing an ultra stiff top. The 18 lbs. of vertical load you are aiming are simply a yoke to copper capability.

I sketched two cross sections of the fins of your block, yust wuess who is more able to bending?

Originally posted by bigben2k
Did your bulletproof glass work OK?

Bulletproof glass is just a very thick policarbonate plate.

[gone_fishin]

Ahh wise advice from nicozeg. KISS is compromised for structural integrity. Who'd have thunk it