> Has anyone considered a new canvas roof
> OVER a new membrane roof? I realize there is
> the added cost, but would there be any
> benefits to having two layers of protection?
> Not being an expert on either and not having
> done either, is there a problem with this,
> i.e. the paint coating for the canvas
> adversely affecting the membrane or
> something similar? Seems like one might be
> able to get the best of both worlds and
> double the protection for the life span of
> the canvas. Or maybe not...

I'd say no. Ordinarily the canvas is applied over tongue and groove siding which offers plenty of ventilation, so the underside of the canvas can "breathe". If the canvas were laid over a rubber membrane, there would be no ventilation, and any moisture that got trapped beneath the canvas due to small leaks would just sit there and cause the canvas to rot out rapidly. So putting on a torch-down roof would seem to be sort of a one-way street.

> One other question. I am currently working
> on the restoration of Buffalo, Rochester
> & Pittsburgh caboose #280. It is a wood
> over a steel frame caboose and has a 1x4
> tongue & groove roof which has a slight
> curve on it as expected. Thus, the boards
> don't fit perfectly together. Plus, we have
> replaced the wood on one end of the roof,
> but the cupola and other end are original.
> The new roof surface is pretty smooth,
> however, the older ones are not as smooth.
> Would there be any benefit to covering the
> T&G with a thin plywood to create a
> smoother surface to then cover with the new
> roof which was canvas and may be again.
> However, we are also considering a membrane
> roof as well. The plywood would be an
> underlayment basically, would not be seen,
> but might provide a smoother surface to roof
> over.

Canvas should be able to ride over small bumps without trouble. Obviously you don't want nails sticking up, sharp corners or shakes, but T&G over a curved roof should be fine. Just run a belt sander over it to be sure. It's not obvious from the photos, but you can see the individual boards through the canvas on the lower part of a clerestory roof. I think you'll never get the canvas so tight that it starts to rip.

OK, why didn't my picture work? Here's the URL for what I wanted to show:

http://206.103.49.193/irm/htm/union205.htm

> OK, why didn't my picture work?

FWIW I've had the same problem on this forum.

mike211@yknet.ca

> FWIW I've had the same problem on this
> forum.

The reason in this case was that you were trying to put an HTML page into the post instead of a JPEG. The image feature on the Interchange will only accept JPEG's (or maybe GIF's, but I haven't tried this), and if you enter a URL with an "htm" or "html" ending it won't work. Instead, I would suggest going to the HTML page you want, right-clicking on the image, selecting "properties," copying the actual JPEG address and putting that in your post. This works using IE, as seen below. Also, using the preview feature is always a good idea.

Frank Hicks

Preserved North American Electric Railway Cars

F-Hicks@wiu.edu

Randall,

Thank you for your response! That is great info and will help me make some better decisions on how to move forward on our roof.

However, I still remain torn (no pun intended) like some others in this thread between original canvas and rubber membrane since the car will not have an indoor storage location for some time. But I now have all the facts which we will have to weigh against each other to come to the best decision for the #280.

One more quick question on canvas while I have you. Since it is painted, have you varied the color of the paint put on different roofs based on original practices as I am sure roofs were painted in different colors. From pictures, it appears the 280 would have had a darker, possibly black roof, although arguably this was probably enhanced by many years of trailing steam engines. Any info would be appreciated. The #280 was built in 1923 by Standard Steel Car in Butler, PA if that makes any difference.

Until later,
Chris


crhauf@frontiernet.net

> One more quick question on canvas while I
> have you. Since it is painted, have you
> varied the color of the paint put on
> different roofs based on original practices
> as I am sure roofs were painted in different
> colors. From pictures, it appears the 280
> would have had a darker, possibly black
> roof, although arguably this was probably
> enhanced by many years of trailing steam
> engines. Any info would be appreciated. The
> #280 was built in 1923 by Standard Steel Car
> in Butler, PA if that makes any difference.

We certainly try to match the canvas paint to the original, depending on the paint scheme. If color photos aren't available (and they aren't always 100% reliable anyway) your best bet is to try to recover some paint fragments from non-canvas parts of the roof, such as the tack moldings or other parts.

Bob Kutella knows a lot more about cabooses than I do, I'll defer to him.

> Randall,

> Thank you for your response! That is great
> info and will help me make some better
> decisions on how to move forward on our
> roof.

> However, I still remain torn (no pun
> intended) like some others in this thread
> between original canvas and rubber membrane
> since the car will not have an indoor
> storage location for some time. But I now
> have all the facts which we will have to
> weigh against each other to come to the best
> decision for the #280.

Chris,

One of the things to keep in mind here is that this is a caboose and not a passenger car or interurban. While cabooses may have come brand new from the builder with painted canvas roofs, I think there is more than ample evidence that the railroads tended to maintain these cars with tar. I've seen more than a few caboose and locomotive cab roofs that were tarred canvas, the fact that tar isn't good for canvas notwithstanding.

Given that your project is a caboose, you have several options. Cheapest would be to just use 90 pound "pilot roofing"; roll roofing without the granuals. Applied on a warm day, this material will fold over the corner of the tack molding, and with a normal nailing pattern of carpet tacks, will look pretty much like canvas from the ground. The biggest drawback is that there will be several lap seams the length of the roof that will need to be tacked to a roof board which really wasn't intended for that purpose. Be sure to trowel a thin layer of roofing cement between the sheets at the lap to glue them together. As the roof ages, it can be coated with asphalt roof cement, just like the prototypes were.

Torch down roofing should also work; my biggest fear is that it will be hard to form over the edge onto the tack molding, and will crack at the corner in time. When it's time to change it, likely all the roof boards will need to be replaced.

Rubber roofing is the "cream de la cream". It can be tacked using a normal tacking pattern on the tack molding, and on a roof as flat as a caboose will lookjust like canvas, tarred or otherwise. The biggest problem with rubber roofing is it doesn't take paint well, but this shouldn't be too much of a problem on a caboose.

Since your caboose is going to live outside, I would recommend against canvas. There is no real advantage,; the authenticity should be hidden under a coat of tar, and unless the car spends most of its days with the stove going inside to keep it dried out, canvas won't have any longer life than tar paper would.

The most important thing to watch for is that the base of the cupola is properly flashed and that the flashing is well sealed to both roof and wall. The best roofing isn't going to give decent service if water can migrate in at the caupola wall. This flashing was often sheet copper, soldered at the corners; often was later replaced by galvanized steel. Avoid aluminum, as it can't be soldered and will corrode quickly anywhere it contacts steel fasteners.

Good luck with the project.

Dennis Storzek

dstorzek@elnet.com

Dennis & Randall,

Thanks again for your very thoughtful and useful responses.

The #280 was heavily rebuilt by the B&O in 1968 which took over the BR&P in the 1930's. Thus the painted & tarred (Aluminum roofing cement) canvas roof we removed was probably from that '68 rebuild. It was an interesting lap seam canvas roof and I must admit, it had lasted quite well given it was 30+ years old when we removed it. However, I think we will go for the membrane roof based on the info you have supplied me.

I will post pictures when we get to the roof some time this summer. I am currently working on the interior appointments.

Until later,
Chris


crhauf@frontiernet.net

> The most important thing to watch for is
> that the base of the cupola is properly
> flashed and that the flashing is well sealed
> to both roof and wall. The best roofing
> isn't going to give decent service if water
> can migrate in at the caupola wall. This
> flashing was often sheet copper, soldered at
> the corners; often was later replaced by
> galvanized steel. Avoid aluminum, as it
> can't be soldered and will corrode quickly
> anywhere it contacts steel fasteners.
> Dennis Storzek

For the flashing to do it's job effectively it must be under the cupola siding and over the roofing material.

Mike

mike211@yknet.ca

> Bob Kutella knows a lot more about cabooses
> than I do, I'll defer to him.

Our supplier makes the canvas paint to order and he has been able to match any color we have given him, even a close representation of a silver metallic. Just in passing, I have seen old print ads extolling the use of canvas for roofing homes! While these were not mansions, the roof was certainly outdoors all the time so presumably they held up well with regular maintenance.

That is probably the key word, since any roof should be inspected and touched up every few years - it is just that few get around to it. Even a '25 year' asphalt shingle roof on your house does not mean you can wait 25 years before you go up there and see how it is.

Bob Kutella


68trolley@comcast.net

This has been a very interesting discussion. It comes just as I was studying the possibility of applying rubber membrane roofs to seven of our steel heavyweight passenger cars which must live outdoors. Our CNR combine 7195 will likely be the "guinea pig" for rubber roofing, with the rest being done over several years.

One other, an 1896 wooden business car, CP 23, is in a million pieces during a long term "as time permits" rebuild. It will receive a pproper canvas roof, as it will always be stored indoors.

Would it be possible for some of the gents who have used rubber membrane (not torch-down) roofing to post information as to sourcing, possible problem areas to watch for, etc.? This would be extremely helpful! Thanks for any info you can provide!

Hmmm, sounds like a good topic for an article! Has anybody written a procedures manual for applying these materials to car roofs?

Steve Hunter,
Cardinal, Ontario


http://www.magma.ca/~sfrm/
s.c.hunter@sympatico.ca

> For the flashing to do it's job effectively
> it must be under the cupola siding and over
> the roofing material.

> Mike

Mike,

Thanks! That will not be a problem for us as the siding is currently off the front and rear of the cupola where the roof meets the cupola.

I do have one other question. Should the roof be adhered to bare wood or painted wood? The recommendation has been made to belt sand the surface smooth, however, should some type of paint be applied? I assume the bare wood has the benefit of better adhesion and no issues with a chemical reaction between the membrane and the coating. And as we discovered, the roof was not painted under the original canvas roof so we are currently starting with bare wood and can go either way. The roof boards will be painted on the underside as they are exposed and form the ceiling of the inside of the caboose.

Until later,
Chris


crhauf@frontiernet.net

> Mike,

> Thanks! That will not be a problem for us as
> the siding is currently off the front and
> rear of the cupola where the roof meets the
> cupola.

> I do have one other question. Should the
> roof be adhered to bare wood or painted
> wood? The recommendation has been made to
> belt sand the surface smooth, however,
> should some type of paint be applied? I
> assume the bare wood has the benefit of
> better adhesion and no issues with a
> chemical reaction between the membrane and
> the coating. And as we discovered, the roof
> was not painted under the original canvas
> roof so we are currently starting with bare
> wood and can go either way. The roof boards
> will be painted on the underside as they are
> exposed and form the ceiling of the inside
> of the caboose.

> Until later,
> Chris

Ours was sanded/ bare wood, as that was recommended by a couple of different individuals and organizations for its adhesion quality. If interested, I have before and after .jpgs I be happy to send via email.

TJG

TJG


Port Huron Museum

tjgaffney@phmuseum.org

> Would it be possible for some of the gents
> who have used rubber membrane (not
> torch-down) roofing to post information as
> to sourcing, possible problem areas to watch
> for, etc.? This would be extremely helpful!
> Thanks for any info you can provide!
> Steve Hunter,

The "rubber" membrane used for roofing isn't rubber but rather a specilaized synthetic called EDPM. It should be available from any quality roofing contractor or wholesaler.
Mike

mike211@yknet.ca

> Ours was sanded/ bare wood, as that was
> recommended by a couple of different
> individuals and organizations for its
> adhesion quality. If interested, I have
> before and after .jpgs I be happy to send
> via email.

> TJG

Speaking about adheasion, it sounds like we are talking a glue-down type of installation. I personally don't think this is a good idea, other than possibly a band of the approved adheasive around the preimiter of the roof. Unlike buildings, railroad cars tend to "weave", that is the structure flexes and the supposedly rectangular roof becomes a parallelogram, distorted first in one direction, then the other. The car building industry found it needed to address this problem when it first tried to fit light steel roofs to wooden or composite freight cars. The result was a variety of "pivoted" metal roofs, where the panels were free to slide slightly to re-adjust themselves to the changing shape of the roof. The problem didn't go away until the advent of "solid steel" roofs, which were strong enough to prevent the weaving by their own stiffness.

The traditional membrane type car roof, painted canvas, resists tearing from this weaving motion due to the fact that it is only fastened at the perimeter, and is simply streatched across the roof like a drumhead. If the material were to be glued to every board, then as the boards slide relative to each other, all this motion would need to be accomodated by a very narrow band of the membrane, and it will likely fail prematurely from the shear. This is a very real possibility with "torch down" asphalt systems, and is likely why some roofers suggest adhearing them to paper rather than the roof. This might not be as much of a problem with rubber, due to its greater flexibility, but why concentrate all the stress in such narrow areas? The only arguement for glueing the membrane down is so it won't blow away in the wind, but it can't do this unless it is ripped, and a ripped roofing membrane is already useless, as it is allowing water into the roof structure. So long as a one piece membrane is firmly attached at its edges, it will be fine. Glueing it across the field could very likely shorten its life.

Dennis Storzek


dstorzek@elnet.com