Sealing Case Halves Ray Greenwood Style!
Dec 7, 2014 19:09:43 GMT -5
Post by vanapplebomb on Dec 7, 2014 19:09:43 GMT -5
This is a writeup that Ray Greenwood posted to TheSamba a while back about sealing case halves. RTV applied by squeezing a thin bead along the faces and smearing with a finger has been a big nono because it can set up too thick between the halves even when torqued down. Here Ray shows an alternative method of applying RTV silicone to case halves so that it goes on very thin and has next to no squeeze out. This is pretty dang cool. I'm tempted to try this method myself. In the last quote there is a link to his Photobucket. Neat stuff. I have a lot of respect for Ray. Check it out!
...
I have used peramtex copper and grey on numerous engines. Never got any in the oil (everyones big...and never leaked a drop.
Its how its applied...not what its consistency it is. If its thick, you squeeze it out on glass or plexiglass and roll it with a hard rubber roller (about $6 ) until it shears out. It will become a uniform thickness according to what the shear characteristics are.
For example permatex copper shears out to a bit under .001" (about 15 microns to be exact). And the gray a little thicker at 18 microns.
Roll your roller around the case parting line to apply it. Reload the roller on the glass. Its fast, clean...and no thicker (probably thinner)...than any of the other sealants mentioned....and works much better in oily enviroments and good to 600F...and stays pliable for eons.
For those confused there are 25.4 microns per .001". So at a hair above .0005" thick... there is nothing to squeeze out and make a mess. All you are trying to do is fill surface imperfections. If the sealer is not transparent after rolling it out (no matter what type it is)....its not thin enough. I can roll permatex black out until its a thin gray transparent film.
You can use this method with any sealer to make a much more uniform joint on case halves. Ray
...
Actually “thickness” is a bad word on my part. It’s a poor description. Bear in mind that though we think of these sealants as a “paste” they are actually a thick liquid.
As a liquid……just like brake fluid or water…..these sealer pastes do not compress. It’s hydraulic.
You can easily have a .001-003” thick layer of sealer “propping” the case halves apart. This happens because the sealer may be so thick that the excess sealer cannot flow out fast enough as you are tightening the case halves.
The wider areas of the case edge are where this will happen first Are you practiced enough with a calibrated eye…. to tell by look and feel how thick of a layer sealant you have after you do this…let it cure…then pull it apart to inspect?
The two main reasons why everyone believes its bad to use RTV on engines cases are
(1) Because they can end up being applied too thick like I noted above.
(2) Because if they end up too thick, invariably excess will be squeezed out into the crankcase and can get into places where it will cause damage (like in a bearing or in a lifter oiling hole).
This is a sad state of affairs…because some of the high performance RTV’s are far better sealers than any of the semi-hardening urethane based aviation sealers also mentioned in this thread……but they must be APPLIED in a totally different manner in order to keep those two points I mentioned above from happening….because the density of most RTV’s…and density is the correct word….not thickness….is orders of magnitudes higher than products like gasgacinch and hylomar.
They are difficult to apply and torque down to get proper thickness in the joint.
For one of the main reasons mentioned in this thread…..the expansion, contraction and minute movement of case halves….RTV like the Permatex Ultra series (and there are many others in the industrial world)…..would be vastly superior (if you could apply them correctly). They have about 300% elongation characteristics without tearing or coming loose.
They are good to about 600F, designed for use in oil…and is fuel resistant.
The nice/lucky thing about RTV’s….is that they are technically “thixotropic”. A thixotropic substance is one that when they are sheared by mixing or carding, they either get thicker or thinner. One that gets thicker is actually called “rheopectic”.
Luckily, “most” RTV’s are Non-Newtonian in nature and reduce viscosity as you shear mix them. You can do this as I mentioned …by rolling them out rapidly on a flat surface with a hard roller. What this does is to reduce the viscosity (temporarily) as the paste shear-thins….by about 30%. It will literally thin the sealer until its transparent.
In minutes...at rest... it will not only regain its original viscosity….it will increase to a higher viscosity as it starts to cure in the presence of oxygen and moisture in the air. So if you roll it out fast and apply it quick…..you can get it onto the case sealing surfaces in a thinner layer than you could even get with hylomar.
You want just enough thickness to fill a handful of microns …which is the void that will always be between the case halves in the form of scratches and the high points of machine marks.
Typically a machined and lapped case half’s edge will have about an 8-10 micron profile on each side. Maybe a bit more or less.
25.4 microns is .001”. So if you can roll out the RTV to about .001” or slightly less and apply it. It will seal perfectly as it compresses into the surface grain of each side of the case and there will be nothing to squeeze out.
I am sealing a case this weekend. Would you like me to send you some pictures? Ray
Its how its applied...not what its consistency it is. If its thick, you squeeze it out on glass or plexiglass and roll it with a hard rubber roller (about $6 ) until it shears out. It will become a uniform thickness according to what the shear characteristics are.
For example permatex copper shears out to a bit under .001" (about 15 microns to be exact). And the gray a little thicker at 18 microns.
Roll your roller around the case parting line to apply it. Reload the roller on the glass. Its fast, clean...and no thicker (probably thinner)...than any of the other sealants mentioned....and works much better in oily enviroments and good to 600F...and stays pliable for eons.
For those confused there are 25.4 microns per .001". So at a hair above .0005" thick... there is nothing to squeeze out and make a mess. All you are trying to do is fill surface imperfections. If the sealer is not transparent after rolling it out (no matter what type it is)....its not thin enough. I can roll permatex black out until its a thin gray transparent film.
You can use this method with any sealer to make a much more uniform joint on case halves. Ray
...
Actually “thickness” is a bad word on my part. It’s a poor description. Bear in mind that though we think of these sealants as a “paste” they are actually a thick liquid.
As a liquid……just like brake fluid or water…..these sealer pastes do not compress. It’s hydraulic.
You can easily have a .001-003” thick layer of sealer “propping” the case halves apart. This happens because the sealer may be so thick that the excess sealer cannot flow out fast enough as you are tightening the case halves.
The wider areas of the case edge are where this will happen first Are you practiced enough with a calibrated eye…. to tell by look and feel how thick of a layer sealant you have after you do this…let it cure…then pull it apart to inspect?
The two main reasons why everyone believes its bad to use RTV on engines cases are
(1) Because they can end up being applied too thick like I noted above.
(2) Because if they end up too thick, invariably excess will be squeezed out into the crankcase and can get into places where it will cause damage (like in a bearing or in a lifter oiling hole).
This is a sad state of affairs…because some of the high performance RTV’s are far better sealers than any of the semi-hardening urethane based aviation sealers also mentioned in this thread……but they must be APPLIED in a totally different manner in order to keep those two points I mentioned above from happening….because the density of most RTV’s…and density is the correct word….not thickness….is orders of magnitudes higher than products like gasgacinch and hylomar.
They are difficult to apply and torque down to get proper thickness in the joint.
For one of the main reasons mentioned in this thread…..the expansion, contraction and minute movement of case halves….RTV like the Permatex Ultra series (and there are many others in the industrial world)…..would be vastly superior (if you could apply them correctly). They have about 300% elongation characteristics without tearing or coming loose.
They are good to about 600F, designed for use in oil…and is fuel resistant.
The nice/lucky thing about RTV’s….is that they are technically “thixotropic”. A thixotropic substance is one that when they are sheared by mixing or carding, they either get thicker or thinner. One that gets thicker is actually called “rheopectic”.
Luckily, “most” RTV’s are Non-Newtonian in nature and reduce viscosity as you shear mix them. You can do this as I mentioned …by rolling them out rapidly on a flat surface with a hard roller. What this does is to reduce the viscosity (temporarily) as the paste shear-thins….by about 30%. It will literally thin the sealer until its transparent.
In minutes...at rest... it will not only regain its original viscosity….it will increase to a higher viscosity as it starts to cure in the presence of oxygen and moisture in the air. So if you roll it out fast and apply it quick…..you can get it onto the case sealing surfaces in a thinner layer than you could even get with hylomar.
You want just enough thickness to fill a handful of microns …which is the void that will always be between the case halves in the form of scratches and the high points of machine marks.
Typically a machined and lapped case half’s edge will have about an 8-10 micron profile on each side. Maybe a bit more or less.
25.4 microns is .001”. So if you can roll out the RTV to about .001” or slightly less and apply it. It will seal perfectly as it compresses into the surface grain of each side of the case and there will be nothing to squeeze out.
I am sealing a case this weekend. Would you like me to send you some pictures? Ray
...
Ok, so here are some basics of what and why I think this is a good method for applying RTV to critical locations like the case halves or to the mating surfaces of transmission case halves (which is where I really started using this method).
Yes there is a link for pictures here down below
My brother, who is not only a mechanical engineer, but also has been building circle track, top fuel, some grand national and NAASCAR engines over the years was the first one to turn me on to proper application of some of these fantastic RTV products that are out there….that many people run away from for anything other than valve covers…because of the worry of getting them into the wrong places or creating tolerance problems. Its an important worry.....but its all in the application.
There are hundreds of very high performance RTV’s out there other than Permatex…many with even better specs..…but the Permatex ultra family is very good as it is……and plenty good for anything we need….and the others may or may not have formulation issues for use around automotive stuff…..why experiment?
The point is, that when I was being shown how to properly apply RTV to a fine tolerance mating surface like an engine case, it was done with the finger in a “stipple work” fashion. If you work diligently and carefully….you end up with an very, very uniform and thin coating with a uniform texture. The uniform texture is key as well. It was a few years before I had enough knowledge of rheology and shear to udnerstand why his method worked so cosnsistently and uniformly.
What I am doing is just a refinement of the stipple method I learned through my line of work.
In general on a VW case edge that is clean, typically the surface machine imperfections are under .001”. They measure out to be (aside from scratches) about 12-20 microns (there are 25.4 microns per .001”). So what you are looking to make is not a “gasket”…but a layer that has enough volume to it to squeeze into and fill the surface imperfections…with only a handful of microns left in between….like maybe 5 microns or less.
Also the thinner you can make the sealing material the less volume there is to flex and stretch..against things like case pressure. In this way…pressurized oil and gasses cannot “tunnel” or make holes through teh stretchy RTV or between the RTV and the case…because the layer of RTV presents so little surface area on its edge…as compared to what it has along its length applied to the case.
How I use this method:
Typically all of the Permatex Ultra series RTV’s rolled out with a hard rubber roller (called a brayer) to about .0018”…1.8 mils.
But...when you apply or roll it to the case edge....The roller only transfers about .0003” to .0005” to the surface of the case edge.
Also…this stuff is so thin it dries to the touch in about 10 minutes or less. It does not fully cure until normal amount of time. On transmission case edges I roll it out on ONE the case edge as thin as possible and let it dry. I roll the other case edge right before I contact the two parts and then let it cure fully.
On magnesium cases like the 004 transmission in my car ..since magnesium is more porous with a rougher surface…I roll both surfaces and then let them dry.
Each surface will have about .0005” of RTV on it…….and about half of that thickness as you can see from the photos…is texture peaks that will crush down. So actual thickness is about .00025 of RTV on each surface...max.
I roll a thin layer on one of the pre-RTV’d surfaces right before I contact them and bolt it up. It never leaks now even with synthetic gear oil.
On the engine case…I rolled out two thin surfaces. I know …because I measured a few times doing this…with the gauge you can see in the photos…..that the RTV layer on a smooth case edge rolls out to about .0003”…with about half of that being crushable texture. So the base layer on each case half…is .00015” with a texture mesh area of about the same.
In general…just letting both of these dry…and then assembling the guts…with no added RTV…makes a bulletproof, leak proof joint. There is “0” RTV to crush out because its dry before you assemble…and the roller gets none on the inside.
Just for notes…I need to buy a new narrow roller (about $10) I had one that is 1” wide that allows me to get around all of the studs and the dowel pins. I lost it somewhere.
It took 2 minutes to roll out the RTV to make it thin on the glass…and 30 seconds to apply what you see here to the case.
This case is not ready to button up so I did not do it all.
To clean the roller up, let it dry fully. Then wash it in hot soap and water and let it dry again. The RTV rubs right off with your thumb in about a minute.
The photos: s1186.photobucket.com/albums/z368/raygreenwood/RTV%20application/
(1) Case half
(2) Glass plate cleaned with acetone. I put a sharpy marker “X” on the back so you can see how thin the RTV rolls out in the pictures
(3) RTV on Glass
(4) Starting roll…roll both directions to get it spread
(5) Starting roll…roll both directions to get it spread
(6) Then just roll one direction to produce a smooth texture
(7) Measuring RTV thickness (which you should not have to do…I’m just doing this to show you how it works)
(8) These are the special gauges I borrowed from the lab at work. These are about $600 each…and they are accurate to 1 micron. They have three precision etched “rails” Two rails are to allow the gauge to roll. The third rail at the edge is chemically machined in a perfect helix….that drops one micron per specific degree sector of diameter. You roll it through the ink, coating, glue..whatever…..and then read directly from the scale on the edge….where it quits contacting the third rail…and this is how many mils (this scale) or microns….on the other…your coating is.
(9) Gauge detail
(10) Case edge detail on bottom of pan
(11) Back of case detail. The black X was made with a sharpy marker first so you can see how even and thin the RTV is.
(12) Texture detail close up
If you are interested I will show you where to buy the rollers on line. Ray
Yes there is a link for pictures here down below
My brother, who is not only a mechanical engineer, but also has been building circle track, top fuel, some grand national and NAASCAR engines over the years was the first one to turn me on to proper application of some of these fantastic RTV products that are out there….that many people run away from for anything other than valve covers…because of the worry of getting them into the wrong places or creating tolerance problems. Its an important worry.....but its all in the application.
There are hundreds of very high performance RTV’s out there other than Permatex…many with even better specs..…but the Permatex ultra family is very good as it is……and plenty good for anything we need….and the others may or may not have formulation issues for use around automotive stuff…..why experiment?
The point is, that when I was being shown how to properly apply RTV to a fine tolerance mating surface like an engine case, it was done with the finger in a “stipple work” fashion. If you work diligently and carefully….you end up with an very, very uniform and thin coating with a uniform texture. The uniform texture is key as well. It was a few years before I had enough knowledge of rheology and shear to udnerstand why his method worked so cosnsistently and uniformly.
What I am doing is just a refinement of the stipple method I learned through my line of work.
In general on a VW case edge that is clean, typically the surface machine imperfections are under .001”. They measure out to be (aside from scratches) about 12-20 microns (there are 25.4 microns per .001”). So what you are looking to make is not a “gasket”…but a layer that has enough volume to it to squeeze into and fill the surface imperfections…with only a handful of microns left in between….like maybe 5 microns or less.
Also the thinner you can make the sealing material the less volume there is to flex and stretch..against things like case pressure. In this way…pressurized oil and gasses cannot “tunnel” or make holes through teh stretchy RTV or between the RTV and the case…because the layer of RTV presents so little surface area on its edge…as compared to what it has along its length applied to the case.
How I use this method:
Typically all of the Permatex Ultra series RTV’s rolled out with a hard rubber roller (called a brayer) to about .0018”…1.8 mils.
But...when you apply or roll it to the case edge....The roller only transfers about .0003” to .0005” to the surface of the case edge.
Also…this stuff is so thin it dries to the touch in about 10 minutes or less. It does not fully cure until normal amount of time. On transmission case edges I roll it out on ONE the case edge as thin as possible and let it dry. I roll the other case edge right before I contact the two parts and then let it cure fully.
On magnesium cases like the 004 transmission in my car ..since magnesium is more porous with a rougher surface…I roll both surfaces and then let them dry.
Each surface will have about .0005” of RTV on it…….and about half of that thickness as you can see from the photos…is texture peaks that will crush down. So actual thickness is about .00025 of RTV on each surface...max.
I roll a thin layer on one of the pre-RTV’d surfaces right before I contact them and bolt it up. It never leaks now even with synthetic gear oil.
On the engine case…I rolled out two thin surfaces. I know …because I measured a few times doing this…with the gauge you can see in the photos…..that the RTV layer on a smooth case edge rolls out to about .0003”…with about half of that being crushable texture. So the base layer on each case half…is .00015” with a texture mesh area of about the same.
In general…just letting both of these dry…and then assembling the guts…with no added RTV…makes a bulletproof, leak proof joint. There is “0” RTV to crush out because its dry before you assemble…and the roller gets none on the inside.
Just for notes…I need to buy a new narrow roller (about $10) I had one that is 1” wide that allows me to get around all of the studs and the dowel pins. I lost it somewhere.
It took 2 minutes to roll out the RTV to make it thin on the glass…and 30 seconds to apply what you see here to the case.
This case is not ready to button up so I did not do it all.
To clean the roller up, let it dry fully. Then wash it in hot soap and water and let it dry again. The RTV rubs right off with your thumb in about a minute.
The photos: s1186.photobucket.com/albums/z368/raygreenwood/RTV%20application/
(1) Case half
(2) Glass plate cleaned with acetone. I put a sharpy marker “X” on the back so you can see how thin the RTV rolls out in the pictures
(3) RTV on Glass
(4) Starting roll…roll both directions to get it spread
(5) Starting roll…roll both directions to get it spread
(6) Then just roll one direction to produce a smooth texture
(7) Measuring RTV thickness (which you should not have to do…I’m just doing this to show you how it works)
(8) These are the special gauges I borrowed from the lab at work. These are about $600 each…and they are accurate to 1 micron. They have three precision etched “rails” Two rails are to allow the gauge to roll. The third rail at the edge is chemically machined in a perfect helix….that drops one micron per specific degree sector of diameter. You roll it through the ink, coating, glue..whatever…..and then read directly from the scale on the edge….where it quits contacting the third rail…and this is how many mils (this scale) or microns….on the other…your coating is.
(9) Gauge detail
(10) Case edge detail on bottom of pan
(11) Back of case detail. The black X was made with a sharpy marker first so you can see how even and thin the RTV is.
(12) Texture detail close up
If you are interested I will show you where to buy the rollers on line. Ray
