How was this made? (Aluminum Heat Sink - Mystery Manufacturing Method)

[Leav]

Ripped this heat sink from an pc, and I am stumped as to the manufacutring method used to make it.

First here's the full gallery.


now to individual descriptions:

General View
Shiny Side
Dull Side - Showing aluminum grains
Single fin - shiny side
Single fin - dull side
Side View 1
Side View 2


The mystery to me is all about the individual fins: Why and how do they all have one dull, and one shiny side?

How was this made? is the shiny/dull intentional or a byproduct of the manufacturing process?

The side views are included as they give a clue (I think): they show that at the bottom all fins have slight radius to one side.
Also, the top of each fin has a noticeable edge.

A lot of people have a lot of experience here on chiefdelphi, so I was hoping someone could bring me in on the secret!

Thanks!
-Leav

Edit:
The heat sink has lots of fins broken off because I was.... investigating them...

[Tom Ore]

Could it have been die cast?

[Raul]

It is hard to tell because the pictures are not crisp enough. But looking at the fact that the fins are so close together and they have no draft and they have a shiny and dull side, I would guess it is a bonded fin process.

http://www.aavidthermalloy.com/produ...in/index.shtml

[JamesCH95]

Originally Posted by Tom Ore

Could it have been die cast?

The aspect ratio on the fins is too big for it to be cast, though the texture on the surface is like that of some cast parts.

I have seen heat-sinks that are extruded (probably not this one), some have the fins bonded into the base. It is hard to tell from the pictures, can you try using a macro setting to get a clear picture of the fin root?

Edit: Raul beat me to it!

[Leav]

Originally Posted by JamesCH95

It is hard to tell from the pictures, can you try using a macro setting to get a clear picture of the fin root?

I couldn't since it's a camera phone, so I built a MacroScope from an old CD drive lens...
(Yes those are my LCD screen's RGB pixels I shot with it!)
Demonstration of magnification:
Outer CD case showing many small dimples.
Single dimple!

Also I cobbled together a work light.

It was impossible to stay focused on the details, so I had to make a video of the area you requested.

If that's not clear enough I could try to make a makeshift stand for it maybe? but with the limited depth of field (I think that's the term) with the CD lens, you can only see a very specific distance each time...


--------------------------------------------------------------------------------------
Back on topic:
--------------------------------------------------------------------------------------
If the fins were bonded on, how would that explain one side being smooth and the other not?

-Leav

[DonRotolo]

I also think it's a bonded process.

What's messing you up is the rough side vs the smooth side, and this may be due to a completely different process issue, relating to how the fins are made (not how they are attached to the body).

How they are made: I am speculating that they are made by a sputtering type process, where the metal is melted and sprayed (as a liquid) onto a flat plate. The back is as smooth as the plate, the front is rough. This is very very fast when forming thin plates and gives a different strength profile compared to rolled-out plates of metal. The rough surface somewhat enhances the heat radiation since it has a greater surface area.

You need the small curve to cover the slot in the mold that the fins fit in to, so the base 'slab' can be poured over it all. If it was straight, it would either need to fit very closely or you'd have to allow some metal to dribble down into the fin cavity. Speculation again, though.

Yes, depth of field is the term. To increase the depth of field, you use a smaller aperture (on a camera, F16 versus F5.6). So, try putting a metal plate with a 1mm (more or less - experiment!) hole in it in front of (against) the lens. You will need more light for an equivalent brightness, but the tradeoff is a greater depth of field. 0.1 mm would not be too small, but lighting some objects may be tricky.

[Big Ideas]

It's from a PC so sputtering process is out, too expensive. Not bonded, too expensive.

The base part is an aluminum extrusion to the outside profile. The gaps between fins are formed in one of 2 ways most likely.

1) there is a gang saw with many saw blades, with spacers in between, on a horizontal shaft that is mounted on a horizontal mill type machine. The roughness would be a function of shoving the parts thru at a high rate. Since aluminum recycles readily, it's better to have a fast process with lots of scrap then a slow process with little scrap. Remember, must be cheep but effective.

2) (this is the fun option) I have seen heatsinks that are formed by a knife that cuts into a solid block of of aluminum in progressive cuts. Just like whittling a fire starter. But this gives slightly curved fins with the root with a sharp angle (not flat).

[Big Ideas]

Looking at the pictures again, I go with the knife (shaving) option. Shiny side is the "cut" side. Dull side is the "ripped" side. The fins are set even, vertical, straight, as a separate operation. This is a low waste and fast process. It has to be a big fun machine to watch. AAvid, and other engineered extrusion heatsink companies do not use this process.

[Tristan Lall]

As mentioned above, it's almost certainly a skived (shaved) fin design.

(You can tell by the slight bend at the base of the fin and the smooth and torn surfaces.)

[DonRotolo]

Yes, I'd buy into that, moreso than a sputtering/bonding process, for reasons elaborated by Big Ideas.

[Greg McKaskle]

Since you are into building cool optical tools, may I suggest building one which will let you view it through a polarized filter illuminated with polarized light. I believe this referred to as polarized microscopy.

I haven't build one myself, but I suggest it because the shiny side in the second photo seems to have circular striations. To me that looks like it is due to cooling on a casting. There are also a distinctive dark mark at the root which looks like a surface abrasion due to forceful clamping or extraction of some supporting surface.

Cool riddle.

Greg McKaskle

[Brandon Holley]

Originally Posted by Tristan Lall

As mentioned above, it's almost certainly a skived (shaved) fin design.

(You can tell by the slight bend at the base of the fin and the smooth and torn surfaces.)

Reiterating this as definitely being a skived fin design. The tell tale sign as Tristan stated is the radius at one side with of the base of the fin.

-Brando

[Tom Ore]

I did a quick search on the internet and found this list:

http://www.tomshardware.com/reviews/...ch,516-60.html

Not sure if this is typical, but the majority appear to be cast.

Here is a picture of one list as cast aluminum:

http://www.tomshardware.com/reviews/...ch,516-10.html

[Brandon Holley]

Originally Posted by Tom Ore

Here is a picture of one list as cast aluminum:

http://www.tomshardware.com/reviews/...ch,516-10.html

If you look at the base of the fins you can see a nice round radius all the way throughout the gap between fins. The heatsink the OP posted about has a very particular shape at the base of each fin where one side is radiused while the other side has a sharp corner. This is a very important detail about how the part was made, because those sharp corners could not have come out of a casting (easily).

-Brando

[Greg McKaskle]

A google search of

images of skived fin heat sink

produces parts that clearly meet what I'd expect from skiving, and do not look like the original part, at least not to me.

The rough side, and what looks like flakes of oxidized scale on the "Single fin - dull side", leave me in the cast camp. I still don't know how they cast it unless they cast the individual fins and bonded them to the base plate. It kind of looks like that in the movie, but of course that is what I'm looking for, and at times, there is enough blurryness to see the Mona Lisa. I'll ask some of the experts at work for other opinions.

Greg McKaskle