Thanks for your technical know-how, folks. In my OP, I never imagined so many detailed responses–or scenarios. Just to clarify: When I mentioned “H-Bomb Samsonite luggage,” I was playing off the idea of a “nuclear suitcase” and not suggesting anything to do with an actual airport. I think my “spinner wheels” quip sent things in that direction. 
Link with pictures of Special Forces parachuting with the “suitcase nuke” ; Special Atomic Demolition Munition, or SADM. More info on development, potential employment and other nuclear demolition munitions.
The SADM was mentioned in the OP, and repeatedly throughout the thread. 
And? I am aware of that. I worked with practice weapons early in my career. The acronym and expansion were included in the blurb from the article.
I added the article as more information related to development and employment. Also pictures.
The lead picture of @smithsb’s interesting cite does give us a very graphic example of how big the thing wasn’t. And that was taken with the device inside its protective carrying case.
A pure fission weapon can be pretty small. I don’t know if there’s some engineering breakthrough that would allow you to miniaturize further than that, but we have been to the point for some time that one strong person can carry such a device.
I don’t know if there’s room for comparable engineering improvements for a Teller-Ulam fusion device. If there were, it would be so classified that it couldn’t come out in public anyway.
The history is fascinating. Fissionable material in sufficient quantity was exorbitantly costly, and in extremely short supply after years of laborious separation from uranium, so much so that Leslie Groves, overall director of the Manhattan Project, ordered a huge (and expensive) steel containment vessel be constructed to capture the material should the first test at Alamagordo be a “dud”. It was never used.
The second device, a “gun type” bomb, was considered so reliable a design that the engineers didn’t even consider testing a requirement. They knew it would work. I think that’s remarkable.
So there are technical or physical considerations on how small a nuclear chain reaction can be? No bombs the size of a shoebox?
You’d do well to read the thread first and also the various wikis and other cites on on atomic and thermonuclear weapons. Which are two very different things.
For an atomic fission device, there are lower limits to the amount of nuclear material required. Smaller than that just won’t go boom. Somewhat bigger than that will be required to solve problems of inefficiency, impurity, etc., of the necessarily less-than-ideal actual physical machine and materials.
As discussed upthread, very small atomic fission weapons can be built. The size of backpacks, but not the size of grapefruits.
This whole thread is a discussion of the smallest thermonuclear = fusion weapon. Which also has an absolute lower limit based on nuclear critical mass, but is overall a much more complicated machine with larger limits all around, lotta magic parts, etc.
Admittedly, I didn’t read everything. I denounce myself.
Sorry, I probably came off shorter than I meant.
LOL, no no no . . . I really should have read in more detail. I knew there was a lower limit on size for nuclear weapons.
Not least because, I think, a fusion weapon actually requires a fission weapon to make it work. Basically, wrap an atomic bomb (fission) around the stuff you want to fuse so it becomes a fusion weapon. Just seems by its very nature it has to be bigger (not to mention more complex).
More complex, but not bigger for a given yield.
That’s fair, assuming “given yield” is “fusion-level yield”.
The largest non-fusion US nuclear test was Ivy King, at 500 kilotons.
It was a stupendously large device, even though it was designed and deployed as an actual aerial bomb. It weighed in at 6,800 pounds and was 5 feet in diameter and more than 10 feet long. (Per List of All U.S. Nuclear Weapons)
The thermonuclear W88 warhead used in the Trident II Submarine-Launched Ballistic Missile has a yield close to this: 475 kT. It weighs less than 800 pounds and less than 2 feet in diameter (at its broadest – it’s a cone, so its minimum diameter is zero). And less than 6 feet long. And the actual fusion device inside that conical reentry vehicle is somewhat smaller than that.
But barring some breakthrough in fusion weapon technology, you’ll probably never get a small thermonuclear device with sub-kiloton yields, simply because currently you need a fission device to trigger the fusion, so that sets the minimum yield baseline.