Paul Valente
Automotive Engineer
In my observations, liquid gaskets take a fairly regular beating by xweb members which fuels misconceptions that liquid gaskets are somehow improper or a lesser gasket than a hard (paper/rubber) gasket. That's unfortunate as liquid gaskets can be perfectly valid solutions and are in some ways superior to hard gaskets. For one thing, they allow metal to metal contact which ensures correct bolt tension throughout the life of the assembly and eliminates the need for re-torquing as paper and rubber gaskets can creep over time. It will also fill-in imperfections in the mating surface to a much greater extent than hard gaskets, will tolerate more joint movement and will seal difficult areas like T-joints. Perhaps a little information will make people not so mistrusting of it.
OEMs have widely embraced liquid gaskets in many engine and drive train applications with great success. If you ever watch the “How It’s Made” TV show when they show the car factories, you’ll see liquid gaskets being applied robotically to all sorts of flanges.
The problems seem to stem from either improper application or inappropriate applications. By that I mean people either do not adequately prep or apply the product wrong or they are applying it to a joint that was not designed to be sealed with a liquid gasket.
I didn’t invent any of this. The chemical companies figured all this out long ago. As I am suggesting you take some of the other forum members posts with a grain of salt, I would suggest you do the same for mine as well. You probably don’t know my background or mental state so by all means, check it out yourself. Here is a very good summary on liquid gaskets from Henkel (they make Loctite products)
http://henkeladhesivesna.com/knowle.../10/14145_gasketing_design_guide-final2-1.pdf
For the sake of this post, I’ll focus on RTV as that seems to be the “goo” that people have the most negativity towards….
Really, for an RTV joint to work well it has to be designed to be an RTV joint. I’m currently rebuilding a spare transmission for my Lancia Beta that has a joint that was designed for RTV so I’ll use that as an example. The Beta transmission is similar to an X1/9 unit in some ways but the Lancia has a joint that runs right down the center of the differential bearing bores. There are two seal carriers that bolt on from either side that seal via an o-ring in a triangular gland. This joint presents some challenges.
1. In order to maintain roundness and concentricity, the cover and case were machined when they were bolted and doweled together. Putting a paper gasket in between them after that fact would make the bores into ovals.
2. Where the seal carriers meet the case/cover there is a T-joint.
You need metal-to-metal contact and you need to seal a T-joint. The Lancia guys figured this out, of course, and designed the joint to be sealed with RTV. How exactly can you spot a joint designed for RTV? Well, you first have to understand a little bit about RTV and I won’t bother regurgitating what you can read on Henkel’s website but basically, RTV that is zero inches thick does not seal very well. Any movement will shear the molecules. So you want some thickness of RTV. What manufactures do is put in features that allow metal-to-metal contact but allow for a continuous bead of RTV. You’ll see steps or grooves sometimes, but the issue with those is that they have a fixed thickness and as it turns out, if the RTV is too thick, it can blow out. So what you need is a chamfer on one half and an overhang on the other.
Here is the Lancia transmission case.
Do you see the 30° chamfer all around the inside of the flange? That is not there by accident! It is so that when the RTV squeezes out from between the flanges, it fills up that triangular space creating a continuous, triangular bead of RTV all around the inside of the joint. If some movement occurs (thermal, stress due to load, whatever) it may locally tear the RTV where it is really thin but the thicker section will maintain the seal.
As (bad) luck would have it, after I sealed this up I realized that I forgot to put the debris magnet in so I had to pull it apart. The good news is, it makes for some nice pictures.
You can see how the flat part of the diff cover flange overhung the chamfer on the carrier.
You can also see that in-between the flanges there is essentially no RTV. It almost all gets squeezed out. I think that is probably the thing that throws most people off is that they see the bead and think that it is just messy and start picking at it. Just leave it!
I went a little heavy on the application. Being it is a transmission, if a piece did break off, and start floating around; it isn’t going to hurt anything.
Here is magnet back in its happy home. You can see there is no chamfer on this side of the joint. There is a bit of a corner round but you can see the ghosted image of where the edge of the chamfer is on the mating part.
So that is what a proper joint that is designed for RTV looks like. There are other factors of course, mainly to do with surface finish and making sure the bolt tension is distributed over your sealing surface, but we can’t influence those at this point so refer to the Henkel document referenced above for more of an explanation if you are curious. Most joints that people struggle applying RTV to on the X1/9 are not designed for RTV in the first place. Also, Fiat seems to have missed it on the bolt spacing and placement in some cases like the infamous cambox. Again, refer to the Henkel document if you want to see where the bolts should ideally be and compare that to actual cambox. It is a lot to ask of any gasket to seal that joint let alone one that it isn’t designed for it.
X1/9 cambox (picture from JJay’s eBay auction). Note there are no features on either part to allow any thickness of RTV to build. No chamfer, groove or step.
Cylinder head is the same thing. Just a sharp corner. This is NOT a good application for a liquid gasket.
The other part of getting a good seal is the application of the product itself. Cleanliness is key! Any oil of dirt will be detrimental to the integrity of the seal. In a production environment you’d check the surface tension of the parts with a special pen (Dyne pen) or you would do a plasma treatment before applying the liquid if you really wanted a strong bond. I've seen Honda air-cooled engines where the valve cover is just glued on (with Hondabond..which is an RTV). No screws or clips. You basically have to destroy the valve cover to get it apart. Since we aren’t going to do either of those, just clean the flanges really well. There are specialty products for this, but brake cleaner ought to do it. Don’t use anything mineral oil based. Make sure you clean the chamfer area too now that you know how important it is. Once it is clean, apply about a 2.5mm continuous bead along the inside about 1mm from the edge and bolt it together. That’s all there is to it.
In conclusion, liquid gaskets are a perfectly acceptable and in some way superior sealing solutions when applied correctly. Don't be afraid of using them in appropriately designed joints on properly prepped surfaces.
PV
OEMs have widely embraced liquid gaskets in many engine and drive train applications with great success. If you ever watch the “How It’s Made” TV show when they show the car factories, you’ll see liquid gaskets being applied robotically to all sorts of flanges.
The problems seem to stem from either improper application or inappropriate applications. By that I mean people either do not adequately prep or apply the product wrong or they are applying it to a joint that was not designed to be sealed with a liquid gasket.
I didn’t invent any of this. The chemical companies figured all this out long ago. As I am suggesting you take some of the other forum members posts with a grain of salt, I would suggest you do the same for mine as well. You probably don’t know my background or mental state so by all means, check it out yourself. Here is a very good summary on liquid gaskets from Henkel (they make Loctite products)
http://henkeladhesivesna.com/knowle.../10/14145_gasketing_design_guide-final2-1.pdf
For the sake of this post, I’ll focus on RTV as that seems to be the “goo” that people have the most negativity towards….
Really, for an RTV joint to work well it has to be designed to be an RTV joint. I’m currently rebuilding a spare transmission for my Lancia Beta that has a joint that was designed for RTV so I’ll use that as an example. The Beta transmission is similar to an X1/9 unit in some ways but the Lancia has a joint that runs right down the center of the differential bearing bores. There are two seal carriers that bolt on from either side that seal via an o-ring in a triangular gland. This joint presents some challenges.
1. In order to maintain roundness and concentricity, the cover and case were machined when they were bolted and doweled together. Putting a paper gasket in between them after that fact would make the bores into ovals.
2. Where the seal carriers meet the case/cover there is a T-joint.
You need metal-to-metal contact and you need to seal a T-joint. The Lancia guys figured this out, of course, and designed the joint to be sealed with RTV. How exactly can you spot a joint designed for RTV? Well, you first have to understand a little bit about RTV and I won’t bother regurgitating what you can read on Henkel’s website but basically, RTV that is zero inches thick does not seal very well. Any movement will shear the molecules. So you want some thickness of RTV. What manufactures do is put in features that allow metal-to-metal contact but allow for a continuous bead of RTV. You’ll see steps or grooves sometimes, but the issue with those is that they have a fixed thickness and as it turns out, if the RTV is too thick, it can blow out. So what you need is a chamfer on one half and an overhang on the other.
Here is the Lancia transmission case.
Do you see the 30° chamfer all around the inside of the flange? That is not there by accident! It is so that when the RTV squeezes out from between the flanges, it fills up that triangular space creating a continuous, triangular bead of RTV all around the inside of the joint. If some movement occurs (thermal, stress due to load, whatever) it may locally tear the RTV where it is really thin but the thicker section will maintain the seal.
As (bad) luck would have it, after I sealed this up I realized that I forgot to put the debris magnet in so I had to pull it apart. The good news is, it makes for some nice pictures.
You can see how the flat part of the diff cover flange overhung the chamfer on the carrier.
You can also see that in-between the flanges there is essentially no RTV. It almost all gets squeezed out. I think that is probably the thing that throws most people off is that they see the bead and think that it is just messy and start picking at it. Just leave it!
I went a little heavy on the application. Being it is a transmission, if a piece did break off, and start floating around; it isn’t going to hurt anything.
Here is magnet back in its happy home. You can see there is no chamfer on this side of the joint. There is a bit of a corner round but you can see the ghosted image of where the edge of the chamfer is on the mating part.
So that is what a proper joint that is designed for RTV looks like. There are other factors of course, mainly to do with surface finish and making sure the bolt tension is distributed over your sealing surface, but we can’t influence those at this point so refer to the Henkel document referenced above for more of an explanation if you are curious. Most joints that people struggle applying RTV to on the X1/9 are not designed for RTV in the first place. Also, Fiat seems to have missed it on the bolt spacing and placement in some cases like the infamous cambox. Again, refer to the Henkel document if you want to see where the bolts should ideally be and compare that to actual cambox. It is a lot to ask of any gasket to seal that joint let alone one that it isn’t designed for it.
X1/9 cambox (picture from JJay’s eBay auction). Note there are no features on either part to allow any thickness of RTV to build. No chamfer, groove or step.
Cylinder head is the same thing. Just a sharp corner. This is NOT a good application for a liquid gasket.
The other part of getting a good seal is the application of the product itself. Cleanliness is key! Any oil of dirt will be detrimental to the integrity of the seal. In a production environment you’d check the surface tension of the parts with a special pen (Dyne pen) or you would do a plasma treatment before applying the liquid if you really wanted a strong bond. I've seen Honda air-cooled engines where the valve cover is just glued on (with Hondabond..which is an RTV). No screws or clips. You basically have to destroy the valve cover to get it apart. Since we aren’t going to do either of those, just clean the flanges really well. There are specialty products for this, but brake cleaner ought to do it. Don’t use anything mineral oil based. Make sure you clean the chamfer area too now that you know how important it is. Once it is clean, apply about a 2.5mm continuous bead along the inside about 1mm from the edge and bolt it together. That’s all there is to it.
In conclusion, liquid gaskets are a perfectly acceptable and in some way superior sealing solutions when applied correctly. Don't be afraid of using them in appropriately designed joints on properly prepped surfaces.
PV
Last edited: