HOW TO JOIN STAINLESS STEEL TO GLASS WITHOUT LEAKAGE OF WATER?

RAHUL JADHAV You would be best served by purchasing a sight glass that is premounted in a threaded housing. Cut threads in your housing and screw in the sight glass. These are available through many industrial supply houses and can handle high pressures if need be. If you do make one from glass you need to have it tempered to prevent it from shattering. Put safety above all else. J. Conway

Here's another method:

But again, take heed of the Safety warning ...

Hello John

Thanks for valuable feedback. But the gap between Glass & Body will create bubble (i observed it in our existing design). Since light is going to pass through this hole i have to keep hole to the rubber gasket. As per our observations, i only feel that if this rubber gasket can fix to the stainless steel body and Glass to the Rubber Gasket & stainless steel body (at cylindrical surface) with some sort of Fast Curing adhesive my problem will be solved.

Thanks & Regards

Rahul Jadhav

You don't specify the pressure or temperature and they're critical. But, I would question the use of an adhesive together with the rubber. What sort of rubber do you have and are you compressing it? If I were doing this with adhesive, I would eliminate the rubber and then call my local Loc-Tite rep to provide the correct adhesive.

Thread the outer portion of the sight glass opening and use spanner nut to apply pressure to your present seal.

Your picture shows no means of how the glass is held in place other than the adhesive. What line pressure is the fluid under? Also what temperature variations are there during operation? If the fluid is heating and cooling the different expansion rate for the dissimilar materials would cause failure of an adhesive.

This could be sealed with out adhesives with some type of retainer to apply pressure on the rubber seal between the sight glass and the block.

Rahul,

This is a classic problem. If you use a flat seal like you show, fluid will leak past it by finding an indirect path.

The solution is to have your sealing occur along a line and not over a wide area. To "float" the glass you will want a rubber gasket of wide area on both sides of the glass. This will allow it to leak because the fluid will find a path regardless of how hard you clamp it with an outer retainer as suggested by others. A gasket is needed on the outside because you don't want to load glass directly or it will break. (every window in houses does this...)

To solve this problem, the gasket needs a "bump" similar to a half o-ring on BOTH sides of the gasket, not necessarily of the same diameter. You could put an O-ring in a groove machined in metal for the between-metal-gasket interface, but the gasket-glass cannot have an o-ring because the glass would have to have a grove made in it which would require a molding of the glass. You could mill a grove in the gasket, but the o-ring will likely "walk" over time.

If you mold a bump into the gasket on BOTH sides of the gasket, then no grove is needed on the metal or glass (lower volume cost). This would require a mold for the gasket.

The key is to have a seal on a LINE and NOT on a WIDE area. This way you will need no glue or bonding (?) of something...just rubber.

Seaplaneguy

I think this is better solution.

But the constraints are i can not change the shape & size of the glass i.e. no other operation can be done it is Optical Glass.

for better understanding u can see new detail image in my previous reply to Mr. Nigh

Design constraints are more

1. Glass size and shape cannot change.

2. Out side dimensions of stainless steel body cannot change. no retainer part can add on the outer surface since this whole body goes into housing.

Regards

Rahul Jadhav

"my previous reply to Mr. Nigh"

What reply? I can't see one.

i am sorry. it was mr. John.

I'm just being a bit pedantic here, but, what you've drawn (simplified and 2D cross section) is this:-

Whereas what you meant to draw was this:-

Yes?

actually it is like this

You could emulate the design of a standard sight glass like this but it looks like you are trying to create a sight tube like this. You could just bolt this on to the end face of your block instead of cross drilling unless you need to be able to look through the tube length-ways for some reason.

If your drawing is representative for the part you have a concept error:

the 2 vertical holes intersect with the sealing surface which generates a gap at the outer diameter of the seal and it is quite normal it will leak.

A flat seal seals when it is in contact all over with the flat surface on which it is applied. It is not the case in your design as it appears.

Correct your design and it will not any more leak provided that you assure enough pressure between seal and contact surfaces and enough radial length of the seal.

I agree with the first half of your post: that's what I was trying to say at post #8, and, his reply at post #10 does not address the perceived problem.

I would have agreed with the second half of your post until I read those of JohnDG (#2) and seaplaneguy (#7). I think they probably understand this sort of thing far better than I do.

Try a 100% silicon rubber adhesive.

http://materials.globalspec.com/Industrial-Directory/silicon_rubber_adhesive

Dear Rohul,

Most leakages where mechanical strength or opacity is not an issue can be easily resolved using EVA adhesive, available universally with applicator. My personal experience using the adhesive is very good to prevent ingress of water when applied on clean surface of most materials.

Good luck

I'm not as convinced as other posters that you can fit these glasses using adhesive. If you can't add mechanical fixings I feel that most adhesives will not bond well enough to a smooth glassy surface to give a reliable long term leak proof joint.

If you do try adhesives could you etch or blast the glass over the bond area to improve the adhesive strength?

I agree, but the OP sounds adamant in his use of an adhesive. He would probably be best served by contacting a adhesive manufacturer for a recommendation.

RAHUL JADHAV, something you haven't told us is the operating pressure & temperature.

Do you really need this big stainless steel block if I were you I would be looking for 2 stainless steel pipe with a ''u'' curve glass pipe since you don't seem to put much pressure glass pipe would be way easier to work with.

Rahul Jadhav - interesting concept seen with the reply from #19 (Yesyen) in this thread. This enables the module to be serviced and repaired or replaced without fear of damage to anything.

However if this sight glass never needs to be serviced or removed again there are products that when used in your application will never leak - but will never facilitate removal again either. It is marketed as Sikaflex and a bead in the corner of the stainless housing with the glass pressed lightly into it will set in about 2hours but will need 24 hours to cure properly.

It'll never leak again, but you will not get it apart again.

Whith many years working with Co2 lasers we had a problem of the resonator tubes becoming exstremly clouded where the cooling water ran through them. We used distilled water with algacid and a couple of other additives and couldn't stop the clouding. If you are going to be running water you may experience the same problem so you may want to design your sight glasses in such a manner as to be able to clean or replace them. Why Is it I think these are not sight glasses but have some other function? J.Conway

google search Sputtering, CVD eqipment. These equipment have glass windows but still achive ultra high vacuum. you will get better idea!

Thanks for the additional info, it puts a whole new perspective on the problem at hand. Properly done the vacuum will increase the sealing forces. Now we need to understand the rest of the dimensions that we are bound by. If dimensions permit make a sleeve that snugly fitts into the clear bore and place an o-ring around the OD of the sleeve. Adjust the sleeve so as to be no higher than the center of the o-ring, you will need to use a size that places the lens above the stainless body so when the body is inserted into the housing it applies sufficient pressure on the o-rings as to create a seal. If the lens is below the surface of the body place an o-ring around the outer surface of the lens to create pressure on assembly when inserted into the housing. J. Conway

Hey, Jerrell, you're replying to kvsubramanyam - his first post in this thread.

Our OP, RAHUL JADHAV, still hasn't replied to my request for more info (the operating pressure & temperature) in my #20 - so we don't know whether we're dealing with a +ve or -ve pressure differential.

I'd assumed (maybe wrongly) that when the OP said " ... But while water is flowing through inlet to outlet it leak through ..." that it implied that water was leaking out, suggesting higher pressure inside the apparatus.

Dear John,

The temperature of the stainless steel body and the liquid is about 37degree. And the pressure is unknown to me. but it is very less. liquid is just flowing from inlet to outlet.

In reply with Comment #12 i could only say that the construction of the vertical hole and the horizontal thru hole is the design requirement. this construction is done to avoid bubbles. In earlier design there was construction like comment #8 image no.2

The current design (which i have posted in #10) is almost freeze after long 6 months trials and improvements and we are very close to the desired results. The only problem we are facing is the leakage. We contacted M/s. Loctite for consultation but they failed to give proper solution. May be i will have to go ahead with suggestion given by seaplaneguy with some other better adhesive.

I am very thankful to others who have participated and spend their valuable time.

Dear Rahul, please comment on my suggestion #19.

Please explain the reason to stick on to gluing only...

Dear Yesyen,

Your suggestion is excellent. But i could not apply it due the size of the stainless steel body ( sq. 12mm) Glass & Rubber gasket (dia 11mm).

regards

Rahul Jadhav

12mm sq, too small a dimension to work with.

I understand your preference for gluing than modifying the basic shape. But once done with due considerations to the design requirement, it is once and for all solved unlike gluing.

Do you have or avail the machining facility for working with this small size?

What is the thickness of the glass? Can a thinner glass be used?

Sorry I did not logged in...

12mm sq, too small a dimension to work with.

I understand your preference for gluing than modifying the basic shape. But once done with due considerations to the design requirement, it is once and for all solved unlike gluing.

Do you have or avail the machining facility for working with this small size?

What is the thickness of the glass? Can a thinner glass be used?

Dear Yesyen

Yes we have the machining facility.

The thickness of the glass is about 1mm. and the rubber gasket is about 0.3mm.

Regards

Rahul Jadhav

If you can get the glass with 0.6/0.5 mm thick, you could recover 0.4/0.5 mm.

And the steel body is to be split into two halves as shown, they can be fastened with set of screws.

A Cylindrical Insert (could be of brass or polymer) has to be machined with needed water passage and this has to be housed between the split halves. An O'ring can be availed to take care of the possible internal leak.

Ultimately the overall diminutions has to remain the same.

Hope rests of the details are self-explanatory or can be discussed if required.

Machine the peice to use a snap ring that holds the glass in tight.

We design/supply purpose-built clarity/turbidity monitors for food industry companies, which comprise a light source (UV led) and receiver (photodiode) either side of a flowing liquid. The assembly requires a leak-tight seal of the window (approx 20mm diam) in the stainless pipe.

In your configuration it would look something like below. It would mean relocating your vertical risers away from the lens but presuming the lens is for viewing the horizontal leg only, this would not seem an issue? For proper sealing this would be a necessary requirement anyway. The seals are material suitable for the liquid and temperature conditions

If the stainless body was machined as per your drawing then creating a seal would be of no problem. The current issue is trying to generate a seal with the vertical bores intersecting the bore for the optical lens. There is nothing to support a gasket or o-ring in that area. The reason for this if I understand correctly is to prevent the presents of an air bubble being trapped in the orginal design behind the lens. I have reasons to believe that the best solution will be to discard the second design body and utilize the first. Please reply and let us know the importance of the optical verses any clear material that will alow the necessary function of this apachure to take place. Can you tell us if this is for visual inspection or some other purpose? If its just for visual inspection there are sight glasses available that seal by means of an O-ring sandwiched between the OD of the the lens and the ID of the bore, some of these are nothing more than a disc with a grove on the OD to contain the o-ring. You may find it possible to make use of this type design now if its not necessary to be of a spicific material. If it is imperative to use your optic then have a grove ground or machined on the OD to acept an o-ring. J.Conway

Since you're gluing. I'm not quite sure what the gasket is for.

Throw away the gasket.

Put lots of glue around the edge of the glass holding depressions.

Clamp the glass discs in place, and stand the block upside down.

Wait to set.

Re-drill the vertical holes (to clear the glue which has blocked them).