Updated Diffusion Pump Assembly and Thermal Modeling

Every fusor and fusion system seems to need a vacuum. This area is for detailed discussion of vacuum systems, materials, gauging, etc. related to fusor or fusion research.
Post Reply
Michael Bretti
Posts: 150
Joined: Tue Aug 01, 2017 4:58 pm
Real name: Michael Bretti

Updated Diffusion Pump Assembly and Thermal Modeling

Post by Michael Bretti » Tue Oct 16, 2018 2:40 am

I have been recently working on the new designs for my vacuum system to incorporate the large 8" baffle to the EO4 diffusion pump, as well as allowing for rapid chamber changes between my three experimental setups. Below is a CAD render of the diffusion pump and baffle assembly with the adapter plates:

Diffusion Pump and Large Baffle Adapter Assembly.jpg

The adapter plates are made from 12"x12"x1" ATP-5 aluminum plate stock, which I will have custom machined to spec when I save up enough. The bottom plate has blind tapped holes on the bottom for the diffusion pump, and an o-ring groove to mate to the bottom of the 8" baffle. The upper plate will have a 6" conflat bolt pattern and knife edge machined into it for adapting to my 6" conflat chambers. Becuase the knife edge is aluminum, I will be using viton o-rings for this mating joint. Long 1/4"-20 bolts will be used through the assembly through 8 holes on the sides to secure it to the 8020 test stand. Below are the CAD renders for the two plates:

Diffusion Pump to 8in Baffle Adapter.jpg
ISO160 Diffusion Pump to 6in CF Knife Edge with Mounting Holes.jpg

Finally a picture and CAD render of the baffle I am using. I obtained this on ebay last year for around $100, and it has been sitting in storage until now.

8in Water Cooled Baffle - Resized.jpg
Large 8in Water Cooled Baffle.jpg

New thermal modeling has also been completed for the above assembly as well, which I will share some info and a project link to the project page in a following post, which includes data, analysis, discussion, and modeling files.

Michael Bretti
Posts: 150
Joined: Tue Aug 01, 2017 4:58 pm
Real name: Michael Bretti

Re: Updated Diffusion Pump Assembly and Thermal Modeling

Post by Michael Bretti » Tue Oct 16, 2018 2:56 am

As mentioned in my above post, I have updated the thermal modeling of the new assembly to better understand how the system will behave thermally, and what to expect during operation for my closed-loop cooling system. A link to the project page is included below:

http://appliedionsystems.com/portfolio/ ... -assembly/

The project page details the scope of the study, as well as improvements from prior studies, limitations, results, analysis, as well as the fusion 360 file for the thermal modeling simulation, and the output report from fusion 360 with all of the modeling parameters. I modeled 6 different cases for the system. The first three look at the heat transfer from the diffusion pump to the baffle and adapter plates during operation for an uncooled baffle for best case, average case, and worst case scenarios. These include: pump cooling of 15C at 25C ambient, pump cooling of 25C at 25C ambient, and pump cooling of 35C at 35C ambient.

The second three cases take the above three cases, and include baffle cooling, with baffle cooling input temperatures at 15C, 25C, and 35C. I won't post all of the pictures from the thermal modeling since there are quite a lot of results, but will include a couple of outputs here for reference. The full details can be found in the link above.

Full thermal scale of the model for an uncooled baffle with diffusion pump cooling at 25C and ambient at 25C:
Full Model thermal Analysis Overview - Cross Sectional View.jpg

Zoomed scale on the full system with baffle cooling at 15C, diffusion pump cooling at 15C, and ambient at 25C:
Full Pumping Stack Zoomed Scale - Cooled Baffle 15C, 25C Ambient, 15C Cooling.jpg

The above model result zoomed in on just the baffle to see the thermal gradients across the component:
8in Baffle Cross Sectional View - Cooled 15C, 25C Ambient, 15C Diffusion Pump Cooling.jpg

Jerry Biehler
Posts: 831
Joined: Tue Nov 24, 2009 8:08 am
Real name:
Location: Beaverton, OR

Re: Updated Diffusion Pump Assembly and Thermal Modeling

Post by Jerry Biehler » Thu Oct 18, 2018 7:39 am

Plates are way thicker than they need to be. You're only talking 15psi here.

I get cast aluminum tooling plate here: http://www.sandsmachine.com/alumweb.htm
Good prices and I have used them for all sorts of vacuum projects.

Michael Bretti
Posts: 150
Joined: Tue Aug 01, 2017 4:58 pm
Real name: Michael Bretti

Re: Updated Diffusion Pump Assembly and Thermal Modeling

Post by Michael Bretti » Thu Oct 18, 2018 12:51 pm

Jerry Biehler,

Thanks for the link, I will definitely keep it in mind for future aluminum tooling plate. I will certainly need more for other designs and projects.

I purchase my aluminum plates from https://www.midweststeelsupply.com/store/, and have already acquired the adapter plates for this design. They offer daily deals and discounts, as well as lots of custom sizes and options. I got these plates at a great discount when I ordered them, and have been very pleased with the service so far.

The thickness of the plates is actually needed for my design, and has nothing to do with holding off atmospheric pressure. Functionally they serve two purposes: to adapt between various components, and for structural mounting. I kept them thick so I can get plenty of thread engagement on the blind tapped mounting holes, especially for the larger 5/8-18 bolts on the bottom for holding the diffusion pump to the first plate.

In addition, the bottom plate will also simultaneously serve as the mounting plate for the entire high vacuum assembly, and will need to be thick and rigid enough to prevent deflection based on how it is mounted, as well as the weight of the entire pumping assembly and chamber, which will weigh quite a lot for the larger ones. This is also why I opted for square stock as opposed to circular, since it would fit my geometry for mounting, and requires less machining.

The top plate is also thick for thread engagement of the blind tapped holes for chamber mounting, as well as for rigidity in clamping both the baffle and the bottom assembly as tightly as possible together without deflection.

A bonus of the thick plates is also large thermal mass and surface area, which helps act as large heat sinks to dissipate thermal energy from the diffusion pump and reduce transfer to the mounted baffle and chamber during operation.

The design will make more sense after you see the full system, with the pump assembly attached to the updated pumping station. I actually just finished integrating the cad models last night and redesigning the foreline to fit the new topology. Once I update the project page on my website, I will post more about it on the forums.

Post Reply