The window is per joint, not per piece
Every data sheet gives an assembly time, and every workshop treats it as a property of the glue-up. It is not. It is a budget, and joints spend it one at a time. A dining chair with eight loose joints, each taking ninety seconds to spread, seat and start closing, needs twelve minutes of budget. A glue offering eight forces the chair into two sub-assemblies before anyone has decided anything.
That is the useful thing about doing the arithmetic first: the split stops being a technique you read about and becomes a number. Two side frames, then the rails joining them, is the standard chair sequence for exactly this reason. It is not tradition, it is the assembly window divided by the per joint time.
The per joint figure is the one people get wrong, and they get it wrong in a predictable direction, because it is usually taken from a dry run. A dry joint slides. A joint with glue in it swells slightly, grabs, and needs the clamp to pull it the last quarter inch, and the rail that went on in twenty seconds during the dry fit takes two minutes with a bead of squeeze-out to work around.
Clamps decide the calendar, glue decides the batches
Two different constraints run this job and they bind at different times. The assembly window decides how many joints go together at once, which sets the number of sub-assemblies. The clamp inventory decides how many sub-assemblies can be sitting at once, which sets the number of cycles, and each cycle costs the full clamp time whether it is full or half empty.
On the defaults here, four chairs of eight joints break into eight sub-assemblies, two at a time through six sets of clamps, which is four cycles and eight hours of clamp occupancy against 7.2 hours of hands-on work. The clamps are the limit, narrowly. Buy three more sets and the cycles drop to three and the bench becomes the limit instead. That crossover is the thing worth knowing before spending money on either clamps or a slower glue.
Cleaning the joint is most of the job
The schedule above is dominated by a number that has nothing to do with glue chemistry: twelve minutes per joint to break it, clean it and dry-fit it. Across four chairs that is 6.4 hours, against 48 minutes of actual spreading. Almost nine tenths of the hands-on time is preparation.
Which is worth knowing before optimising anything else. A glue with twice the assembly window saves you a sub-assembly and changes the total by minutes. Anything that halves the cleaning time changes it by hours. Whether a given joint even can be cleaned without damage, and what the original adhesive was, is a question about the piece rather than about the plan.
Where this sits
Clamping pressure and clamp counts for the glue-up itself are the glue and clamp calculator, and the joints being reglued are usually the mortise and tenon calculator. If the seat comes out at the same time, that is the seat deck calculator or the caning take-off calculator. Refinishing after the frame is solid is the finishing schedule calculator, and whether the whole exercise pays is the furniture refinish breakeven calculator.
Questions people ask
Where do I get the assembly time for my glue?
The technical data sheet, and it will be given at a stated temperature and often a stated humidity, both of which matter enormously. Assembly time and open time are not always the same thing on a data sheet either: one may refer to how long the glue can sit spread before the parts are brought together, and the other to how long the whole operation may take before clamping. Read which one the sheet means. If you are working somewhere warm, the sheet figure is optimistic and your own trial with a scrap joint and a stopwatch is worth more than any published number.
Why round the joints per window down instead of up?
Because a joint that is half closed when the glue grabs is worse than a joint that was never started. Rounding up would tell you to attempt nine joints in a window that holds eight, and the ninth is the one that will not seat. The page rounds down and prints the exact figure beside it so you can see how close the call was. A result of 5.9 rounded to five is worth looking at twice, since a small improvement in the fit might genuinely buy the sixth.
Does clamp time mean the piece is finished?
No, and the two are separate fields here for that reason. Clamp time is how long the parts have to be held before they will stay put; full cure is how long before the joint has whatever strength the manufacturer describes. They can differ by a factor of ten or more, and a chair that comes out of clamps at lunchtime and is sat on that evening has been asked to do something the data sheet did not promise. Both figures come off the sheet and this page repeats whichever ones you enter.
Should I take the whole piece apart or only the loose joints?
That is a judgement about the piece, and there is a genuine argument on both sides that no calculator settles. Taking everything apart means everything is clean and the piece goes back square; taking only the loose joints means fewer joints disturbed but a frame that may not pull together because the sound joints hold it in the shape it has taken. What the page does is show what the count costs: run it once with the loose joints and once with all of them, and the difference in cycles and hours is the price of the decision.
Is old glue in the joint a problem?
It is the main problem, and what to do about it depends on what it is, which usually cannot be determined by looking. Some old adhesives can be softened and cleaned out, some cannot, and a new glue over a poorly cleaned surface is bonding to the old adhesive rather than to wood. Beyond the technical question there are two prior ones on old furniture: finishes and fillers on pre-1980 pieces can contain lead, which makes scraping and sanding a testing matter first, and anything of real value belongs with a conservator before it is taken apart, because the original joinery and the original surface are usually most of what makes it valuable.