I have a setup where I want to move a knife like blade through a body of liquid inside a container and I want to control the blade from outside the container. Apart from using magnets to move the blade, is there any way to form a dynamic seal around the blade as it moves along the container length?

I was thinking of using two long lengths of rubber that sandwiched the 1mm blade, but I am not sure how this would hold once some pressure was added to the container, especially if the blade had an oddly shaped cross section (like sharp curves that would be hard to seal).

There is a fused silica layer at the top of the container, which is in contact with the liquid.

Vat with scraper blade

So I am wondering if there are any seals anyone knows of that could be used in this scenario? I'm thinking of possibly using ferrofluid as a seal at this stage.

  • $\begingroup$ Why a sharp blade? exact shape of movement? What exactly do you want to achieve? $\endgroup$ – mart Jan 15 '16 at 10:00
  • $\begingroup$ I don't think a ferrofluid works the way you think it does. If you're moving this through water, and you think that water is going to touch the ferrofluid, then you're going to contaminate the water with whatever solvent/oil the ferrofluid is made with. Not to mention the fact that ferrofluids only coalesce around magnetic flux lines (not continuous), and they only hold as well as the strength of the magnetic field. Why not just use purge air if you need to prevent water from getting in? I don't know if this will work or not because your problem statement is too vague. $\endgroup$ – Chuck Jan 15 '16 at 20:19
  • $\begingroup$ I just added a picture to help explain my question a bit better. The setup is for use in a UV based resin 3D printer. Basically the large container will be completely filled with a photopolymer resin, up to the quartz glass ceiling shown in the picture. The small white blade is needed to scrape the partially cured resin off the surface of the glass twice a second as UV images are shone through it, which means the blade will have to move through the resin. It will need to be connected to our linear drive on the outside of the container, and hence require some sort of seal. $\endgroup$ – user88720 Jan 16 '16 at 15:10
  • $\begingroup$ @Chuck the ferrofluid I was planning on using is apparently immiscible with the resin and shouldn't mix. I was thinking of using two multipole magnetic strips on each side of the gap in the container that the blade runs through, which will be covered in the ferrofluid. I'm fairly confident it should work when the blade is static, however I'm unsure as to how it will hold up if the blade is moving as high speeds. What are your thoughts? $\endgroup$ – user88720 Jan 16 '16 at 15:36
  • $\begingroup$ My thoughts are why not just make the blade a U shape and put it on the inside of the vessel instead of needing some elaborate seal. Whatever partially cured resin is going to get on the inside U blade will get on the blade regardless. Regarding the ferro fluid, just because it's immiscible doesn't mean it can't contaminate your resin. Solids are immiscible, but I wouldn't want to drink out of a lead cup because it contaminates water. If there is contact, there is the potential for contamination. $\endgroup$ – Chuck Jan 16 '16 at 20:10

Depending on the axes the blade shall be allowed to move, i.e. the degrees of freedom of your blade, you should try to seal the axes separately. A solution would also depend on anything besides the blade being allowed inside the container.

The best solution (seen from sealing) would be if you can live with additional rods inside your container. Then you could construct a system of rods or levers to move the blade around and just seal the rods. There are lots of sealing solutions for moving rods around that hold up to a few thousand bar against liquids.

If that's not possible and the blade's cross section is your only concern, you could add a mechanical adapter to your blade with the blade's cross section inside and some more easy to seal cross section outside as in the following sketch.

enter image description here

If that is not enough, you are looking for something like a linear moving seal. Depending on the pressure difference you want to seal you might go with a system of plates that has tongues and grooves which are made with a transition fit or slight interference fit, e.g. made from teflon or some liquid-repellent material.

  • $\begingroup$ Yes I can change the contour of the 'blade' at the seal, so this may be an option. When you mention the moving rods, do you mean seals would seal against the lateral movement of them, or simply like a cylinder seal? From memory, the first thing I googled was 'linear moving seal'. I didn't get many results so I thought I might have just been using the wrong terminology, hence why I thought I would ask on here if there was such a thing. $\endgroup$ – user88720 Jan 16 '16 at 14:57
  • $\begingroup$ When sealing rods, the usual options are static seals (no motion between the parts) or dynamic seals. Sometimes dynamic seals are further subclassified by rotation-only movement or linear movement. 'linear moving seal' is appropriate. For small pressure differences (few bar) and some leakage allowed, O-rings should be fine. Otherwise there are complex shaped seals (C-shape etc.). The optimum seal depends on many parameters (e.g. movement, pressure difference, materials, temperature, surface roughness). I suppose you can either go trial and error, or you ask somebody who sells seals. $\endgroup$ – Robin Jan 17 '16 at 13:29
  • $\begingroup$ Ah I see you mean rods that would be on the inside of the container right? $\endgroup$ – user88720 Jan 17 '16 at 13:36
  • $\begingroup$ Yes. That's what I meant with 'if you can live with additional rods inside your container' in my answer. If you cannot add the rods directly to the blade, probably you can add some minimum-disturbing support to the edge of the blade where rods are mounted. $\endgroup$ – Robin Jan 18 '16 at 7:19

Directly Coupled Rodless Cylinders used sealing technology that holds up to over 100 psi of air, see Rodless Cylinders

  • $\begingroup$ Interesting read. However when they say "this is manufactured from a high performance resin and is carefully moulded to fit the exact profile of the slot in the cylinder barrel", I get the feeling that it may be a little too complex for my application. $\endgroup$ – user88720 Jan 15 '16 at 9:47
  • $\begingroup$ Building one from scratch may be cost prohibitive, but buying an used cylinder and scrapping it for the seal mechanism might be feasible. $\endgroup$ – Greg Marsh Jan 15 '16 at 18:03
  • $\begingroup$ Hmmm that is true. I'll send them an email and see what I can come up with. Cheers $\endgroup$ – user88720 Jan 16 '16 at 15:03

One option is not to try to design a leak-proof seal, but instead just limit the leaks to acceptable levels and make maintenance possible.

My suggestion:

  1. Use rubber seals like you planned to.
  2. Add a trough that will catch any leaking resin. Preferrably made of a material such as teflon or silicone that the resin will not stick to.
  3. Arrange so that the resin in the trough will eventually harden, making it easy to clean off.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.