Machining and welding custom high vacuum chamber - Page 3
Close
Login to Your Account
Page 3 of 3 FirstFirst 123
Results 41 to 49 of 49
  1. #41
    Join Date
    Dec 2008
    Location
    Vt USA
    Posts
    6,734
    Post Thanks / Like
    Likes (Given)
    736
    Likes (Received)
    2316

    Default

    Quote Originally Posted by rcoope View Post

    SNIP

    I would actually start by buying a text like https://www.amazon.com/Foundations-V.../dp/0471175935 There is a lot of secret principles, eg the efficiency of the mechanical pump lines scales with like the 4th power of diameter and each elbow cuts flow by a factor of two so you can increase your pump down speed by a order of magnitude with just a bit of care. Likewise there's lots of lore about how clean things need to be, venting your screws and screw holes to avoid virtual leaks etc. I found that chamber I linked too above on my first google attempt but that's because I know what I'm looking at. A text book would really help orient all these things.
    I will disagree on the usefulness of large or straight mechanical pump lines as an aid to realized pumping speed. Pumping a small volume such as a 24 inch bell jar or a 30 inch box coating system is fast and easy Regardless of the roughing line size (Within reason! Call it 1.5 inch dia) And in fact, smaller diameters help keep the lines in the viscous regime and inhibits back flow and oil migration.

    It is pumping the SURFACE within the system that takes the time. (Reference post #30 )
    And at 10E-3 and lower, even backing the diffusion pump, any size tube will suffice. (1.5 inch or so) One just must be mindful of backing line pressure when choosing the cross over.
    I've done it both ways, Big pipes, big gate valves, big pumps. In a typical process cycle (batch coating) there is nothing to be gained in going big. And the big lines BEG for oil back flow when roughed for extended periods of time. Small lines are much more forgiving .

    Of course, automated sequencing of valves takes care of the back flow issue. But adds sophistication and expense, both upfront and maintenance.

    Best thing I have used to increase productivity and reliability of diffusion pumped systems is a Dry N2 bleed through the process chamber sufficient to keep the foreline pressure around 75-100 microns during the entire pump cycle. Turn the N2 off if not desired during the process. But back on for "air release". Note! N2 can get expensive if "time is not money". The dry N2 additionally serves as a gas ballast that helps keep the mech. pump oil water free. Money in the bank!

    Considering that vacuum is mostly nothing, there is a lot to it! ;-)

  2. #42
    Join Date
    Aug 2005
    Location
    Raymond , CA
    Posts
    237
    Post Thanks / Like
    Likes (Given)
    0
    Likes (Received)
    147

    Default

    Quote Originally Posted by snowman View Post
    All of them....except platinum, that can be melted in air.

    And yes, it's a temperature range, but the amounts will be small, melted on a liquid cooled copper hearth.
    Snowman are you planing to melt platinum in a electron beam heated copper crucible and then cast the melt into a mold?

    If so I suspect you will need a taller chamber. The height of the top of the hearth will be about 4 inches above the chamber base plate. This is a minimum to fit either base plate mounted or collar mounted electrical, water and mechanical motion feedthroughs. You will also need a view port window mounted on a tube coming off the chamber wall at 45 degrees to allow the melt to be viewed. You may also need a window for a infrared sensor to monitor the melt temperature.

    The electron beam travels in a arc about 1 to 2 inches above the crucible edge before arcing down into the melt. A thermal radiation shield is also needed above the crucible to capture any metal vapor emitted by the melt and to keep the top of the chamber cool. The electron beam arc, view port access and shield requirements will add at least 6 more inches of height required above the crucible.

    Tilting the crucible for a pour will be a challenge with the cooling water connections. The alternatives are either a bottom pour from a fixed position crucible or using a poco graphite crucible liner which can be lifted out for the pour.

  3. #43
    Join Date
    Jul 2004
    Location
    Southeast Michigan
    Posts
    6,545
    Post Thanks / Like
    Likes (Given)
    239
    Likes (Received)
    1450

    Default

    Quote Originally Posted by Robert R View Post
    Snowman are you planing to melt platinum in a electron beam heated copper crucible and then cast the melt into a mold?
    I am not smart enough to build an electron beam furnace. At least not yet, maybe in time....zero exposure.

    This is just a small arc furnace for laboratory sample preparation. We already have a small button melter, but it's vacuum capacity is a bit lacking...so it's more of a fancy tig welder accessory. So the first order of business will be to put together another small button melter that is capable of high vacuum, good measurement, with a mass flow controller to back purge the right gasses back in at a known rate.

    There was an article posted in 2011, where a group made a tilt'n pour using a T section of 250 mm ISO-K, modified with a sight glass as you said. The whole chamber was then tilted, allowing a modified copper hearth to pour into the mold cavity.

    Cooling water is a necessary evil that I'm already accustomed to dealing with.

    So this chamber is related to all of that.

  4. #44
    Join Date
    Aug 2005
    Location
    Raymond , CA
    Posts
    237
    Post Thanks / Like
    Likes (Given)
    0
    Likes (Received)
    147

    Default

    Quote Originally Posted by snowman View Post
    I am not smart enough to build an electron beam furnace. At least not yet, maybe in time....zero exposure.

    .
    There is not much involved. Your lab probably already has a 7 cc to 40 cc capacity electron beam evaporation source and power supply sitting on the shelf. The source is run at low power with a beam sweep to form the melt. The next step would be to increase the power for the evaporation operation. You just need to skip the second step.

    It sounds like you have a miniature vacuum arc remelt furnace. A arc is struck between a mold and a electrode (cathode) made up of the metal to be degassed. The electrode gradually melts drop by drop filling the mold. The contaminants are released as the drop is formed.

  5. Likes snowman liked this post
  6. #45
    Join Date
    Jul 2004
    Location
    Southeast Michigan
    Posts
    6,545
    Post Thanks / Like
    Likes (Given)
    239
    Likes (Received)
    1450

    Default

    Quote Originally Posted by Robert R View Post
    There is not much involved. Your lab probably already has a 7 cc to 40 cc capacity electron beam evaporation source and power supply sitting on the shelf. The source is run at low power with a beam sweep to form the melt. The next step would be to increase the power for the evaporation operation. You just need to skip the second step.

    It sounds like you have a miniature vacuum arc remelt furnace. A arc is struck between a mold and a electrode (cathode) made up of the metal to be degassed. The electrode gradually melts drop by drop filling the mold. The contaminants are released as the drop is formed.
    Well, *my* lab is my garage, so no...we don't. The facility I'm working with only has induction melting capability right now.

    We have a vacuum arc button melter. The electrode is a tungsten rod. While there is some mass transfer from anode to cathode, it's not much. Mostly we are using the heat created by the arc for virgin melt of chemically purified metal sponge. Following virgin melt, you have to remelt multiple times to get rid of oxygen left in the sample.

    This is the diffusion pump rack we purchased. Looks like it has thermocouple gauge, not ion as I thought. So I believe that between that and a valved mcleod I can use them to know when to move rougher from chamber to pump, and turn on diffusion...but it doesn't have a liquid nitrogen trap.

    Is that a conflat between the vacuum takeoff and the pump unit?

    What is with the brass fittings to the right of the pump?

    I may be better off using the pump out of the electron microscope that I have as I grow, but I liked the "modularness" of this unit, simply for learning curve of getting through the basic operation of a diffusion pump.

    I very definitely have a learning curve ahead of me.

    Applied Test Diffusion Pump System | eBay

  7. #46
    Join Date
    Aug 2005
    Location
    Raymond , CA
    Posts
    237
    Post Thanks / Like
    Likes (Given)
    0
    Likes (Received)
    147

    Default

    The pump stand is roughly 1/2 century in age. The plumbing style was used by the Vacuum Research Corporation and later by Varian for their leak detectors.

    The diffusion pump is air cooled with the foreline pressure monitored by a thermocouple gage. The diffusion pump has a butterfly type isolation valve mounted above a adapter spool. There is a second thermocouple gage mounted above the butterfly valve. The brass plumbing mounted to the right probably holds a vacuum operated electrical switch, The four bolt flanges are a ASA standard. There are no conflat seals on the adapter spool.

    The diffusion pump will be filled with a hydrocarbon based fluid rather than a silicon based oil. It will have a high vapor pressure but will form a electrical conductive film when polymerized. The stand was probably used to pump out a electron tube before sealing. The low conductance of the bellows hose determined the stand pumping speed. The system did not work well for its intended application. This setup is what gave diffusion pumps a poor reputation.

    The roughing pump has a missing exhaust port. I do not see a ballast valve.

    The general rule in surplus vacuum equipment sales is to set the asking price to be five times what the item is actually worth to a informed buyer.

    The informed prices are dropping. I was surprised to see a small stainless steel cryo pumped box coater with an electron beam source and associated instrumentation sell for $5000 earlier this year from a NIST lab. It was a internet sale on the Public Surplus website.

  8. #47
    Join Date
    Jul 2004
    Location
    Southeast Michigan
    Posts
    6,545
    Post Thanks / Like
    Likes (Given)
    239
    Likes (Received)
    1450

    Default

    Quote Originally Posted by Robert R View Post
    The system did not work well for its intended application. This setup is what gave diffusion pumps a poor reputation.

    The roughing pump has a missing exhaust port. I do not see a ballast valve.

    The general rule in surplus vacuum equipment sales is to set the asking price to be five times what the item is actually worth to a informed buyer.
    LMFAO!...thank you for your honesty! And your description.

    At least I can't screw it up too much more. I mean...I'm sure I can, but hey...maybe this will be rock bottom.

  9. #48
    Join Date
    Jul 2004
    Location
    Southeast Michigan
    Posts
    6,545
    Post Thanks / Like
    Likes (Given)
    239
    Likes (Received)
    1450

    Default

    I just want to thank everyone who has taken time to respond to this thread. I've got a much better handle on what I'm doing now, and it seems a little less insurmountable.

  10. #49
    Join Date
    Aug 2005
    Location
    Raymond , CA
    Posts
    237
    Post Thanks / Like
    Likes (Given)
    0
    Likes (Received)
    147

    Default

    Hello Snowman:

    I have attached a description of the closest match diffusion pump I could find to the pump that you purchased. I suspect that your diffusion pump will have a 400 watt heater and a 5 cfm 3/4 HP roughing pump if there is a close match.

    A generic 24" diameter bell jar system is often supplied with a 3000 liter/sec air speed high vacuum pump with a chamber designed to reach 1**10-6 torr base pressure. These systems often have stainless steel sheet metal liners to make cleaning easier, extensive fixturing, shutters, heaters and deposition sources. The actual surface area of the system will be at least four times the surface area of the bare chamber.

    If the same bell jar only needed a ultimate vacuum of 1**10-4 torr and everything else remained the same a 30 liter/sec high vacuum pump would be required.

    Vacuum arc remelting is done at 1**10-3 to 1**10-4 torr . A lower vacuum provides no improvement in degassing for the iron and titanium alloys. The same may be true for your platinum sponge.

    Your 300 liter/sec diffusion pump may do the job. What you need to do is cut a hole in the top of the cabinet, remove the diffusion pump hose manifold, and raise the diffusion pump flange level with the top of the cabinet. Your chamber base plate is then bolted directly to the diffusion pump butterfly valve housing.

    The brazed copper foreline and roughing line plumbing is ok for a first attempt at arc melting. If the system works you may then want to upgrade it to the standard KF flange fittings with air operated valves, molecular sieve
    trap, pressure gages, and a automatic valve sequencer. Your initial goal is to melt metal and not be concerned with a ideal vacuum system.
    Attached Thumbnails Attached Thumbnails scan1.jpg   scan10004.jpg  
    Last edited by Robert R; 05-28-2019 at 11:20 PM.


Tags for this Thread

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •