I can see your thinking outside the box, which is a good thing, but I see some concerns in your post. I'll state the first and most obvious one first.
The source cross section/focal size you've quoted implies iridium 192. So I'll go with that assumption for the time being.
D4 film will require somewhere in the neighborhood of 4.5 to 5R to properly darken the film, so for academic purposes, I'll go with the 5R figure.
IR192 outputs 5.9R per hour at 1 foot; you have a 3" tube; you wish to shoot this panaramic; if it's standard size, it should be in the neighborhood of 3.25" but I'll stick to 3" to make it even.
@ 1 foot and 1 curie you'll output 5.9R per hour. @6" it will be 23.6R @3" it will be 94.6R, and @1.5" (where your source would be set for a panaramic) it will output 377.6R per hour or 6.29R per minute. Half value layer of steel is about .5" depending on the steel for Ir192, so you'll have a .5 hvl round about. This will put your shot time somewhere in the range of 1 minute with 1 curie.
At that low a level, you'll have to calc out the "exact" activity of that source, as 1.2ci vs 1, or .8ci vs 1 will have a dramatic effect on shot time, vs say a 10ci source at 12" the difference in .2ci will not be very noticable if noticable at all.
Factoring in the half life of IR192 at approx 74 days, you'll have a very limited window in which your source will be viable for this kind of shot.
From the quality standpoint, you have to remember that source size is over half of your wall thickness and will be shot from a centerline of 1.625" away from the surface of the film assuming intimate contact of part and film. At that distance the difference in diameter of that source will have an effect on your shot, and your UG will vary for the same reason. If I am not mistaken, the standard UG formula does not take into account what you are proposing. It would also have to be remembered that you will have to calc that ug to the surface centerline of the pill it'self at the distances your proposing.
Simply put I don't see this being possible from a quality standpoint or an economic standpoint, at least not with IR192.
Having said all that there is an option of you go with a selenium75 source.
http://www.asnt.org/publications/materialseval/basics/feb99basics/feb99basics.htmIn 2002 I took part in an effort to qualify a mothballed reactor head for service. Some of the original test records could not be found, so the friction welds on the stalks had to be retested. Given the distances between stalks, shooting them with standoff was not a real option. Given that I've sleep many nights since then, I can't recall the exact parameters used on that one, I do recall the answer was selenium75 with shots taken from the ID of the stalk out.
Given that an 80ci Se75 source can be obtained with a .118 cross section/focal size and that our source was special ordered and had a (distance memory so verify for yourself first) .05 cross section at 10ci it worked out for us and the NRC.
.146" will not work, but .05 should. A half life of 74 days will not be economically feasable to maintain, but a hl of 120.4 days would be. 5.9hr/per ci is not feasable either as they are typically sent back long before due to the hl, I doubt You'll find an Ir192 source less than 10ci for that reason (10ci of ir192 would be near impossible to shoot with at those distances due to flashing the film and the super short shot time). Then there is the <=2.8 or so R/h value for Se75. All in all, I believe this is a possible answer for what your intention is. You could special order the source, and with the added half life, keep that source cooking for a year vs 3 months for the ir192. Se75 also works great for CR radiography. That is another option for you if time is your concern.
In summary, what you propose I believe may be possible, but not in the manner initially proposed.
Suggest further research into the matter.
Regards,
Gerald