Ed-
Thank you for your thoughtful and informative reply.
Your comments indicate, to me, the importance of including the application data and allowing for reported and observed
variances in the color over time. A field for this with a suffix to the "official" paint number (if any) should be a part of the
database to allow for this.
Not all fields for every entry will be able to be populated.... your horror example (for us) of a lot of cars to be painted to match
a drift card sample without any brand, number, name of the paint is an example.

I envision doing the database prior to dealing with the actual drift cards, means of reproducing them, etc... and as an ongoing
work.
This database will, incidently, provide some framework of drawings and other documents in existence as a byproduct. While it
will, because of the thrust of the project, be limited to General Arrangement and Lettering & Painting drawings and specifications
that normally contain color information, it would be something.

Charlie Vlk




Charlie,
For the period 1931-1952, the AC&F bills of materials have numerous
paint samples that were affixed to the back covers of these documents.
As many people know, Pat Wider, Ray Long and I researched these BOMs
and received permission to cut a small sliver of the paint sample to
keep. I have a collection of these paint samples. Anyone can see the
samples by visiting the St. Louis Mercantile Library (Barriger National
Railroad Library) and go through the BOMs.

From the standpoint of freight car colors used in this period, in many
cases the BOMs specify a generic name of a paint color without the
mention of a brand of paint. Others specify the brand of paint plus a
name of the paint color (i.e., DuPont Tuffcoat Brown). In very few
cases are there paint numbers that were specified. Not all orders BOMs
had paint samples (absolutely none on tank cars), so there are many
examples where there are no samples available for various road names or
private owners. In other cases there are perhaps up to a dozen paint
samples used on various cars for a given railroad.

In addition to the AC&F material, I have a collection of Pullman BOMs
that apply, for the most part, to cars built at Bessemer, Alabama from
1929 to 1947. Included in this collection are a relatively small number
of BOMs of cars built at Pullman's Butler, Pa. plant. These documents
have no color samples, but they tend to specify the paint brand/paint
name in the same way that the AC&F BOMs specify them. Some have paint
numbers, but most do not.

The AC&F paint samples clearly show that paint colors changed over time
even though the name of the paint may have remained the same. For
example, Santa Fe Mineral Brown changed significantly from the samples
of the late 1930s/early 1940s to those from the late 1940s.

Many of the paint brands and color names have already been published in
RP CYC articles and/or my articles in Railmodel Journal. A main point
that I want to emphasize is that each AC&F paint sample represents the
color applied to one order of cars. The same color may have been used
on other lot numbers that were produced, but in many cases the colors
changed from one order to another even for the same railroad. Sometimes
multiple brands of paint were used for the cars built in one lot
number. In some cases there are multiple paint samples in the BOM and
sometimes there's only one sample. It can get pretty complicated. but I
am willing to be a participant in a study to help define the colors to
the extent that they can be defined. The challenge is to come up with
some way to match these paint samples to some type of standard that's
available for anyone to use.

Of course, these paint samples document selected cars when they were
brand new. Samples for repainted cars typically done at the railroad
shops will have to some from other sources.
Regards,
Ed Hawkins