My point is No mater where you get your so called standard color chip you can not match it.
Let say you want to mix "River Bottom Blue" Get your chip from where ever you want. It calls for a gallon of base 17. You are not sure the base is the same shade as when the chip was made. It calls for 3 Oz of pigment QW. 5 Oz of FG. You start with a clean 3 and 5 Oz measures. You put you pigments into the measures and then pour them into the base. If you look at your measures you will notice some of the pigment remains in the them. You do not know how much so you measurement is not true. Never True. So you guys just keep on mixing paint.
I do not need any right now. 8>)
Say good night "Gracie"


Thank you
Larry Jackman
ljack70117@...
Boca Raton FL 33434
I want to die in my sleep like
my grandfather did, not screaming
like the other people in his car.

Tony -

A toad by any other name is still a toad!!

A.T. Kott, P.E.


--- In STMFC@..., Anthony Thompson <thompson@...> wrote:

A.T. Kott wrote:

Just as an aside, in the late 1950's, some MP gondolas were

repaired

with "Hup-bolts".

I think this product is the "Huck" bolt.

Tony Thompson Editor, Signature Press, Berkeley, CA
2906 Forest Ave., Berkeley, CA 94705 www.signaturepress.com
(510) 540-6538; fax, (510) 540-1937; e-mail, thompson@...
Publishers of books on railroad history

A.T. Kott wrote:

Just as an aside, in the late 1950's, some MP gondolas were repaired with "Hup-bolts".

I think this product is the "Huck" bolt.

Tony Thompson Editor, Signature Press, Berkeley, CA
2906 Forest Ave., Berkeley, CA 94705 www.signaturepress.com
(510) 540-6538; fax, (510) 540-1937; e-mail, thompson@...
Publishers of books on railroad history

Dennis -

Thanks for the great posts on rivet size. I went home and looked in a
few of my files to see what sizes were actually on a couple of cars.
Here are the results:

AT&SF Tk-J tank car shell - 1" OD, flattened on top (not conical,
like on modern tank cars - these were old tanks placed on newer
frames in the teens).

AT&SF Bx-12 SS boxcar - 1" OD rivets, steel on steel; and 3/4" square
bolts, 3/8" thick, with 7/16" OD threaded stub 1/4" long on hat
braces, etc. (wood to steel).

CB&Q SS boxcar - 1" OD rivets, 5/8" square nuts.

C&IM bathtub gondola - 1-1/8" OD rivets.

C&IW composite gondola - 3/4" square nuts.

D&H #19607 DS boxcar - 1-1/8" OD rivets in underframe crossbearers.

MP #86162 50' steel express boxcar - 1-1/8" OD rivets on ends, 5/8"
rivets on vertical side seams, 3/4" rivets on bottom row of rivets on
sides. (Typical steel boxcar of the 1940's).

I-GN #6342 SS boxcar - 3/4" OD rivets on ends; 7/8" square nuts (with
no thread showing) on Zee braces, 5/8" OD rivets on steel sheathed
wood doors.

SP F-50-13 flatcar - 1-1/8" OD rivets on poling pockets, ends; 1-1/4"
OD rivets on sides; 1-3/8" OD rivets on stake pockets; 1-1/8" OD
rivets on crossbearers in underframe; 7/8" OD rivets on bottom angle
of fishbelly centersill.

M-K-T steam loco tender - 7/8" OD rivets.

T&NO 60-B-5 baggage car - 3/4" OD facia rivets, with every 24th rivet
being 1" OD; 3/4" OD rivets on sides, except: 2nd row from bottom has
1" OD rivets and the bottom row has 7/8" OD rivets.

T&NO 60-C-5 coach - all rivets in roof are 1/2" OD; sides all 3/4"
OD, except both bottom rows are 7/8" OD.

All of the above are ACTUAL measured rivet/bolt HEAD dimensions on an
ACTUAL car! There is some confusion in these posts as to whether a
car used 1/2" rivets or the rivets used have a 1/2" OD head. Dennis'
post has the correct information which will allow you to estimate
head size.

Just as an aside, in the late 1950's, some MP gondolas were repaired
with "Hup-bolts". These, for purposes of modeling, appear as 1" OD
cylinders which stick out about 1-1/4" from the car side. Hot
applied rivets were unreliable when you calculate their strength,
since they could be a range of temperatures when applied. After they
were headed over, they shrank, pulling the sheets together, and the
friction between the sheets gave the side of the car its strength.
The Hup-bolt, on the other hand, was a two-piece assembly, consisting
of a cylinder with a flat-head nail through it. The shank of the
nail had a notch in it. These were applied to the car as follows:
(1) The "nail" was run through a hole in the sheets from the inside.
(2) The cylinder was slipped over the shank of the nail on the
outside of the sheets. (3) A machine gripped the cylinder and pulled
the shank of the nail until it broke, simultaneously crushing the
cylinder. In this way, the force between the sheets was well known,
not as variable as with hot rivets. These Hup-bolts are REALLY UGLY
on a car, in my opinon!

Take that, you stinkin' rivet counters!!

A.T. Kott



--- In STMFC@..., "Dennis Storzek" <destorzek@...> wrote:

--- In STMFC@..., "Tom Madden" <tgmadden@> wrote:

Here's a formula from Wayne Long:

The "size" of the rivet (in inches) is the outside diameter of

the

shaft which goes into the two or more holes in the metal pieces
being riveted together. The diameter of the rivet head, where it
bears on a metal surface is size x 1.5 + 1/8". The head at its
highest point off the same metal surface is:
size x 1.5 + 1/8" x 0.425.

Apply this information to a standard 1" rivet and you get:

Shaft = 1" O.D.

head = 1 5/8" O.D.

Tom, List,

This got me curious, so I did some digging. Machinery's Handbook has
the ANSI specifications for solid steel rivets with button heads (at
least my mid-1980's edition has them). Wayne Long's formula is right
on for 1/2" rivets, but gives slightly large values for smaller

rivets

and slightly small values for larger rivets. For instance, a 1" ANSI
button head rivet has a 1 3/4" head diameter. However Wayne's

formula

is certainly close enough for our modeling, as the error in HO scale
would be less than .001". For those who care, here are the sizes of
the heads for ANSI Standard button head rivets, and their HO scale
equivalents:

SIZE HEAD DIA. HO SCALE

1/4 .460" .0055"
3/8 .684" .008"
1/2 .875" .010"
5/8 1.094" .0125"
3/4 1.312" .015"
7/8 1.531" .0175"
1 1.750" .020"

I might point out that .020" dia. is the smallest size separate
styrene rivet that Tichy makes, and it models the head of a 1" rivet
in HO scale.

So, what size was used where? That's a rather had question to

answer,

and took some searching. Typically rivet sizes aren't specified on
general arrangement drawings; one needs to go down to the component
drawings level to find them, and few of these are published. I went
back to that set of AAR drawings from the Field Manual / Interchange
Rules we were discussing the other day, and gleaned size
specifications from the call-outs on the roof, end, and door

drawings.

Here is what I found:

The smallest I found was 1/4", used on the lap eams of the AAR flat
roof; similar to the PRR X29 roof, but note this is NOT the PRR
specification;

The rivets that attach the roof flange to the top of the end are
called out as 3/8";

The rivets that attach the ends of the roof seam caps to the side
plate, and side sheathing to the car posts in the door area are

called

out as 1/2", all the rivets on the sides are drawn similar, but I
could not find a spec. other than at the door posts;

One of the ends called for 11/16" holes along its bottom edge, this
would be the clearance hole for 5/8" rivets;

Striker castings were attached with 3/4" rivets;

The largest rivets I found were from a different source, the ARA
(later AAR) standard for splicing steel center sills, which called

for

7/8" rivets.

Nowhere did I find any 1" rivets specified, indicating that the

Tichy

product is only good for bridge rivets in HO scale.

As an aside, in my design work at Accurail I've always used .010" -
.011" for rivet heads on car sides, slightly larger for frame

rivets,

and these are indeed very close to scale for 1/2" rivet heads. I
suspect any rivets that look much smaller, the designer

misinterpreted

the specified nominal size of the rivet for the actual diameter of

the

head.

Dennis

Strictly speaking, this may be off-topic to Steam Era Freight Cars somewhat, but it seems an important enough general problem
that bears on the purpose of the list.....
There is a Yahoo! Group that hasn't gotten off the ground
rrColorStds@...

that was formed to explore just such issues. For some reason it hasn't gone anywhere.

In our vast pool of talent there have to be some commercial artists, paint chemists, or even color theorists that can jump in and make something like this happen.

I see a number of separate but related activities necessary to make a Standard Railroad Color Reference Library for Model Railroads:

1) A uniform catalog of real railroad paints and their applications. There are some DuPont and other lists available to begin building this database. One activity would be for everyone to report paint references on any prototype sources they have as it appears on the document (not only did the method of listing colors change by the manufacturers, often railroads and builders mis-listed the color numbers or called them by the wrong name. This would have to be recorded and rectified to make a master list that is correct and complete reference). Many colors were reused between different railroads (Santa Fe Red as used on the Warbonnet Fs and the Red used on Burlington Route E Unit stripes, for example). A list of paint colors with what railroads used them on what would be a great framework for developing a Color Library.

This would take the combined efforts of someone who is familiar with data base input on the internet and a group of prototype savy people to suggest at least an initial range of possible inputs that need to be recorded. After this is done a call would go out to RR Historical Societies, modelers, the various forums and email groups on the internet for folks to review their collections of drawings to contribute color reference data along with the full citation of the source.

2) A uniform way of obtaining reproducable copies of official Color Drift Cards. These would be the base starting point for developing Standard Railroad Colors for use on Railroad Models (and restorations of equipment in Railroad Musuems as well). Somebody with work experience in reproducting color in a uniform, reproduceable manner would be invaluable to this effort.

3) Some work needs to be done on ways of adjusting colors from the prototype samples so they look "right" on models under typical layout viewing conditions. Some research has to be done to see if there are any uniform approaches for adjusting colors for scale and reduced lighting typical of most Model Railroad situations. It might be reasonable to make some recommendations for achieving reasonable color balance with commercially available lighting (especially in light (pardon the pun) of the move to energy efficient fluorescent replacements for incandescent bulbs so that the recommended colors can be seen as intended. Probably in conjunction with this work some standards for paint finish (gloss, semi-gloss, eqgshell, flat, textured, etc.) should be done as well with the same considerations for scale and viewing conditions.

4) As part of this work, some comments on real world lighting and how to interpret photographs of real equipment might be further information worth putting into the knowledge base for Model Railroaders.

5) The colors could be referenced in a manner that computer users could reproduce them (within the limitations of their monitors, graphics card and printer devices) for non-model use.

I am sure that I have only scratched the surface of what needs to be done regarding establishing a Standard Railroad Color Reference Library for Model Railroads, but I am sure that others can jump in and help make this project a reality.

I would encourage those interested in this subject to join the rrColorStds@... group as that would be the proper forum for meaningful work towards this end, but it may be that the moderators here will allow a little discussion from those with thoughts on the subject but not interested in joining the grpup per se.

Thank you,

Charlie Vlk

Many years ago, I worked as a color formulator for Mobil Chemical. One job that passed across my lab bench was a color match for Western Maryland covered hoppers.
The WM specified that we match to a Federal Color Standard color panel. These were maintained by the US Navy in Philadelphia and any paint company could obtain uniform color standards from this source upon request. The standards were referenced by a five-digit code number which were often referenced in military, government and industrial paint specifications. The advantage of this system is that it is a tightly controlled single source of closely matched reference panels that can be used as paint color standards by any private or government agency wishing to solicit bids from paint manufacturers who wish to become qualified paint suppliers to the bidding agency.

I have seen a number of railroads use this system but, as we were a supplier of vinyl paints, we only solicited business where certain chemical resistance was required and vinyl paints fit the WM's criteria for covered hoppers. One negative property of vinyl paint for this application was that vinyls are notorious for fast color fading when exposed to sunlight and, as a result, these covered hoppers would start chalking very quickly after being placed in service. This was not a big concern for the WM, however, as this type of paint protected their investment in coveved hoppers very well against chemical and impact damage. Besides, the light gray color masked the chalking very well.

Jim Kubanick
Morgantown, WV

--- In STMFC@..., richtownsend@... wrote:

I have used Grandt Line caboose windows.

You're right Richard. The oners I used were Grandt so maybe the Tichy
ones aren't too big.

Ed

Good points, Tony.

Does anyone have any ideas on developing a "standardized" color/paint
reference that come CLOSE (subjective, of course, and that's been
hashed out here & elsewhere) to the most common car & locomotive
colors used over the years? It would not have to be be exact, as
exact isn't even attainable (as Larry says).

Using common paints? model paints? inks? RGY/CMYK? other??

I've seen numerous model paint mixing tables specifying "X oz.
Floquil Brunswick Orange plus 7 drops reefer black"...etc, but I'm
afraid if commercial paints are used to reference prototype colors
they will change or be discontinued over time. We need something
that will work now and over time.

Nothing is going to be perfect, but in this case, close enough is
close enough. At least we'll have something.

Rob Simpson



--- In STMFC@..., Anthony Thompson <thompson@...> wrote:

Jim Eckman wrote:

I'd love to see a STMFC Panatone list for the more common

railroads,

It would make a great starting point for mixing!

It's "Pantone" and I think you're wrong for two reasons, Jim.
First, many colors fall between Pantone numbers, or outside the

gamut.

It's rare that a Pantone color MATCHES any specific color pulled

out of the real world. Second, it specifies ink mixtures (for
printing), using percentages of the proprietary Pantone basic inks,
and would not be possible to translate to mixing of any particular
paint.

The idea of the Pantone system is that a designer can choose a
color from the array which suits what they want to do, and be

assured that the printed work will NAIL that color. Matching some
particular, pre-existing color is another matter entirely.

It's sometimes true that a color, such as PFE orange, is
fairly close to a Pantone color, and that was the color we used to
print the background of the dust jacket of the PFE book. But many
colors, such as the many "box car reds" of the prototype railroads,
simply do not have close matches.

Tony Thompson Editor, Signature Press, Berkeley, CA
2906 Forest Ave., Berkeley, CA 94705 www.signaturepress.com
(510) 540-6538; fax, (510) 540-1937; e-mail, thompson@...
Publishers of books on railroad history

Jim Eckman wrote:

I'd love to see a STMFC Panatone list for the more common railroads, It would make a great starting point for mixing!

It's "Pantone" and I think you're wrong for two reasons, Jim. First, many colors fall between Pantone numbers, or outside the gamut. It's rare that a Pantone color MATCHES any specific color pulled out of the real world. Second, it specifies ink mixtures (for printing), using percentages of the proprietary Pantone basic inks, and would not be possible to translate to mixing of any particular paint.
The idea of the Pantone system is that a designer can choose a color from the array which suits what they want to do, and be assured that the printed work will NAIL that color. Matching some particular, pre-existing color is another matter entirely.
It's sometimes true that a color, such as PFE orange, is fairly close to a Pantone color, and that was the color we used to print the background of the dust jacket of the PFE book. But many colors, such as the many "box car reds" of the prototype railroads, simply do not have close matches.

Tony Thompson Editor, Signature Press, Berkeley, CA
2906 Forest Ave., Berkeley, CA 94705 www.signaturepress.com
(510) 540-6538; fax, (510) 540-1937; e-mail, thompson@...
Publishers of books on railroad history

Jim-
In my industry experience, Pantone Colors do not come close to most railroad colors.
There are at least 900 distinct DuPont colors (at least I know of that many drift cards!!) and DuPont was only one
of many paint suppliers to the railroad industry.
My guess is that we'd need color matches for around 2000 colors and another 100 or so to duplicate metals, weathering, woods, etc..
It is amazing how far off colors can be even when you have access to common color specification systems from all over the world.....
You can develop mixes using RGB, CMYK etc. proportions, but even there you need further work to control other factors such as brightness,
finish, etc...
The factories always want someone to spec a single number to absolve them of any responsibility for the final color.
It isn't that easy for anyone involved in the process. Only if it were that dialing in a magic number could result in a correct color for every model!
Charlie Vlk
Railroad Model Resources


I'd love to see a STMFC Panatone list for the more common railroads, It
would make a great starting point for mixing!

Jim Eckman

.

I'd love to see a STMFC Panatone list for the more common railroads, It would make a great starting point for mixing!

Jim Eckman

Anthony Thompson <thompson@...> wrote:

>>I'd put the MDC Harriman cars up against any competition for "most

oversize HO rivets."

Those rivets are nearly scale sized compared to Bowser and Mantua tender rivets!

Ray Breyer





---------------------------------
Be a better friend, newshound, and know-it-all with Yahoo! Mobile. Try it now.

The Royal Horticultural Society has a color matching system available that consists of a very large set of standardized color cards; each has a hole in it. One holds the cards close over the item to be matched.

It has the advantage that it can be used in various lights and will provide an exact match tailored to the user. Each card is identified by a color number so one can convey the color match to others using the system.

W. Lindsay Smith wrote:

Professor please! Your discussion of prototype rivets is accurate. However, since the human eye resolves about a mil (1 foot at range 1,000 feet) and the effect of shadows makes small projections appear larger, the model maker cannot make automatic scaling.

It was Kurt Laughlin who provided rivet dimensions, not me. And though I happen to agree that rivets (and other features) may have to be other than scale size to "look right," I didn't discuss that either, in the current thread. Whether you offer praise or blame, Lindsay, it doesn't come to me <g>.

Tony Thompson Editor, Signature Press, Berkeley, CA
2906 Forest Ave., Berkeley, CA 94705 www.signaturepress.com
(510) 540-6538; fax, (510) 540-1937; e-mail, thompson@...
Publishers of books on railroad history

Professor please! Your discussion of prototype rivets is accurate.
However, since the human eye resolves about a mil (1 foot at range
1,000 feet) and the effect of shadows makes small projections appear
larger, the model maker cannot make automatic scaling. The late Dick
Kurtz made the original Milepost 310 trailer rivits to scale and they
did not apear on the model. He redesigned the rivet row and we
thought they looked more like the trailers we saw in TOFC service.
Love ya Tony!
Lindsay
--- In STMFC@..., Anthony Thompson <thompson@...> wrote:

Kurt Laughlin wrote:

Tony, do you think that the "small" rivets (actual nomenclature

for

rivets under 1/2-inch) were driven hot? I would think that by WW

II

if not earlier they would be cold formed. (I have a photo

somewhere

of an M3 tank of 1941 vintage being assembled with an enormous

riveter

and bucking bar. The holes are 49/64 dia -.765 - so I imagine

they

were using 3/4-inch rivets, "large" in the terms of the trade.)

Cold-driven rivets are much stronger, while hot ones hold the

seam

tighter. Depends on what you want. With a tank, I can guess the
priority <g>.
Certainly as late as the 1950s, PFE reefers were still

assembled

with hot rivets. I can't speak for freight car production in

general.

Ed Kaminski might know what AC&F practice was.

Tony Thompson Editor, Signature Press, Berkeley, CA
2906 Forest Ave., Berkeley, CA 94705 www.signaturepress.com
(510) 540-6538; fax, (510) 540-1937; e-mail, thompson@...
Publishers of books on railroad history

Colour blindess--it's more common than you think. There is at least
one well-known model rail author that is colour blind. Colour vision
is tested by either the Isihara plates (those coloured numbers inside
coloured circles) or a lantern test, which originated with the Royal
Navy, but is also used--for testing prototype train crews. AFAIK
there are few colour standards--AAR signal colours, Pantone colour
system, and inks used in printing the Isihara plates.

Then we add those "color guides"--Charlie, you are quite right. The
graphics standard for colour is the Pantone system. I often wonder
about the accuracy of some of those colours in the Morning Sun books
that I enjoy and will buy more of. And there are also some rail
books that have colour digitally altered to improve a photo's
appearance. Some publishers seem to employ this technique often!

The best prototype colour references IMHO are drift cards, wet
samples, or Pantone references. We CN modellers are fourtuante to
have had Stafford Swain and company do up a "chip of many colours"--
essential for CN/GTW/CV/DW&P modellers.

Steve Lucas.

--- In STMFC@..., "Charlie Vlk" <cvlk@...> wrote:

This is a problem that keeps on coming up.

The funny thing about "correct" colors is most of the time the

frame of reference by the vast majority of consumers is

another company's model (correct or not) or "Color Guides" printed

in China from decades old color prints or slides

(or even colorized B&W prints!!!) color matched using Chinese or

Japanese ink systems and printed by people who

never saw the prototype!!

Drift cards are a good start and reproducing these valuable source

documents is a huge problem.

Prototype
Lighting Conditions
Viewing Distance
Sky / Weather Conditions
Sun Angle
Weathering (dirt, etc. film)
Paint Fading / Deterioation
Varying Paint Formulations
Different Vendors
Changes in Specifications over time
Paint Finish

Model
Scale of Color
Lower Intensity (than the Sun) Lighting Conditions on the Layout
Different Temperature Light Sources
Cumulative errors in matching colors from sub-masters
Color Perception of Individuals involved in the

Research/Design/Manufacturing/Approval/Production processes

Color Perception Memory
Scale of Paint (pigment size....evident in Metalics and composites

like Graphites

Paint Finish (largely not addressed in models beyond Gloss/Flat)
Sprayed paints vs. Tampo printed inks
Order of application of paints over varying color substrates

...and a high percentage of Males and even some Females have some

degree of Color Blindness

Charlie Vlk
Railroad Model Resources

[Non-text portions of this message have been removed]

Excuse the tardy reply. Tom Madden's memoire on this subject is extremely interesting, not to mention very informative on this once very controversial subject.

I never met Paul Mallery, but I always have paid close attention to what he said and wrote- even now. He was a very prolific and very influential author in our field for many, many years. His book on modeling bridges is still the very best around, although the advocated methods of modeling contained within have been long superceded.

I love the drawings and diagrams in all of his books, and his explanations of the details of special trackwork is not duplicated elsewhere. I have used copies of his full sized fold-out HO turnout diagrams to gauge tie spacing, etc. and help with some hand-laying chores. His long advocacy of only hand-laying track (with epoxy "spikes") gained no traction whatsoever that I could ever determine, although he advocated it for years on end full in the face of the obvious rising hegemony of sectional flex track, which he disdained.

The weak point of his writing over the years was his quite evident inability to even indirectly admit defeat when points of view that he championed were either over-run by historical events, the powers of empirical observation, the better thinking of others, or just mere logic- and too often that persistent failure to give way gracefully tainted his credibility on other things where he was on stronger and higher ground. He disdained other prominent modelers' reputations and work by the pretty deliberate use of faint praise, and he almost never would acknowledge the work of others if he perceived that their work competed with his own. Both A.C. Kalmbach and Linn Westcott were soft targets in these regards. I always thought that this was peculiar for a man of science nominally devoted to the superiority of reasoning over personalities.

If Paul Mallery ever wrote about Steam Era Freight Cars, I am unaware of it. However, warts and all, and on balance, there is no doubt in my mind the model railroading environment through which our cars run today is measurably better because of Mallery's well publicized efforts.

Sadly, he passed away in retirement four or five years ago in Roseville, CA (quite near here).

Denny
--
Denny S. Anspach, MD
Sacramento

Ed Mines wrote:

I don't want to argue with Dennis Storzak who has done so much for the
hobby but I think .010" is too big for "rivets" on box car sides.
Where does it say that they are "rivets" anyway? Could they be bolts?

Tony Thompson:

In wood-sheathed cars and stock cars, yes, they often are bolts.
So you make a good point, Ed, that modelers sometimes refer to "rivets"
on wood attachments. But on steel car sides, they are rivets. And yes,
they were applied hot.

Do you really need 1/2" diameter rivets to hold sheet metal (or thin
metal plates) to a frame? And so many of them? . . . There are bolts
with 1 inch diameter heads on a railroad underpass I see. There aren't
too many of them either and the service is much more severe than
holding on freight car siding.

Tony Thompson:

Loading may be greater, but a freight car experiences bending
and twisting in a way that bridges do not, so you cannot directly
compare the fastener needs; and the numerous rivets on the car sides
are to seal the seam of the panels, not just to attach them together.
It's probably true that they would be adequately ATTACHED by a rivet
every foot, but the seal would not be very good under that twisting,
etc.

There is another aspect to this besides sealing the panels together against weather. These panels
work together to provide some of the strength of the car body. You have all seen the photos in
which you can see the "oil-canning" of the side panels due to stresses. If you had but 12" spaced
rivets, that would simply come apart at the seams (to coin a phrase). By spacing them much closer,
they make the side panels work together as a larger sheet, as a diaphragm.

SGL

In bridges (many carried steam era freight cars) the rivets were driven hot,
through the 1950's, since the connection design was slip critical.
Brian J Carlson P.E.
Cheektowaga NY

Kurt Laughlin wrote:

Tony, do you think that the "small" rivets (actual nomenclature for rivets under 1/2-inch) were driven hot? I would think that by WW II if not earlier they would be cold formed. (I have a photo somewhere of an M3 tank of 1941 vintage being assembled with an enormous riveter and bucking bar. The holes are 49/64 dia -.765 - so I imagine they were using 3/4-inch rivets, "large" in the terms of the trade.)

Cold-driven rivets are much stronger, while hot ones hold the seam tighter. Depends on what you want. With a tank, I can guess the priority <g>.
Certainly as late as the 1950s, PFE reefers were still assembled with hot rivets. I can't speak for freight car production in general. Ed Kaminski might know what AC&F practice was.

Tony Thompson Editor, Signature Press, Berkeley, CA
2906 Forest Ave., Berkeley, CA 94705 www.signaturepress.com
(510) 540-6538; fax, (510) 540-1937; e-mail, thompson@...
Publishers of books on railroad history