Model accuracy is an ongoing process, not a corrective action after the fact.
The key is understanding SketchUp’s inference system of snaps, dynamic guides and constraints.
• The user controls the drawing axes location and orientation.
• The drawing axes steer the inference engine.
• The inference engine steers the tools.
• Inference locking, via the four Arrow keys and Shift key, eliminates off-axis errors.
• Numeric keyboard entry via the Measurements toolbar insures dimensional accuracy.
Given the model shared by the OP is reasonably close to desired real-world size…
The threaded portion is only ~0.250” long.
A thread die will come up against the underside of the head, long before any finished threads are cut.
Cutting clean threads up to a shoulder would be lathe work.
@hewz , Given the model you shared is reasonably close to desired real-world size…
3D printing diminutive functional threads may ultimately prove expensive and disappointing.
Much depends upon the properties of the specific print material, the resolution of the print technology used for that material and losses due to post-process finishing.
Printed in certain metals, it’s doable, provided one models the mating parts with careful consideration of proper fit, material shrinkage and finishing losses.
See this guide to 3D Print Material Properties and Processes
More info here…
As a learning exercise in precision 3D modeling and printing, it’s a worthwhile endeavor.
As a means of procuring an ordinary thumb screw, 3D printing is a poor choice when such things are mass produced by far more accurate and economical processes.
Have a look over here … seven pages of thumb screws.
That was the conclusion I came to as well. I purchased a tap and die set last night and in no time my parts were assembled. Proof of concept is my focus at this stage and not production. I was able to quickly identify the limitations of using a nut and bolt approach, so, I am implementing a design change to eliminate that.
Geo, thanks for your guidelines on how to model with accuracy in Sketchup and the links and references you supplied as well. I am currently reviewing them.
Based on your recommendation, I purchased a tap and die set last night which proved to be a wise decision because it allowed to quickly determine “proof of concept” and showed me that I needed to change course, thus saving me a lot of time and expense.
Thank you and thanks to all who provided any feedback.
Incidentally, if you are still planning on the screw, it would be wise to design it with a shoulder or shank under the head. As it is, it’s a broken screw waiting to happen due to the stress risers at the head. With the shoulder’shank, a standard die would work just fine for finishing the threads. Besides, if there’s a washer or something else being retained by the screw, you wouldn’t need the thread to run all the way to the head. If you really had to run the threads to the head, a die that size wouldn’t be very expensive. You could get one and grind down the back of it to get rid of the lead-in. Start finishing the threads from the front of the die and then turn the die over to get the last little bit.
No, The actual design (not shown) involves using the screw to support interfacing two circular gears. The central, single point support, is not sufficient to maintain the even distribution of force needed to allow the gears to rotate, relative to each other, consistently.
I initially thought of using the screw, as a means of having the ability to, finely, adjust the force you would need to apply in order to rotate the gears.
The new direction I am going involves using multiple fixed-length tabs, equally spaced around the much larger central core of the circular gears. In essence, I am physically distributing the support more evenly using tabs that will allow me to snap the gear components together. The challenge now is just to determine the length of the fixed tabs and I will use the screw as a guide.
I should probably check my messages more often, sorry for the late response.
The major points have been pretty much covered in the other replies (some really good advice and tips in there)
My original vid is based on real engineering standards/scales so should be able to be finished using a tap and die set.
The biggest issues are the closeness of tolerance in SU vs the real world, the fact that arcs are not true arcs (in the CNC sense) and the quality of your print/printer.
Subtracting the bolt copy from the nut blank leaves exactly 0.0mm clearance. If you could print them exactly to size (and that’s a big IF), the surface of the bolt thread would occupy the same space as the surface of the nut thread. You would need to scale the nut up, or the bolt down, through the X/Y axis only so as to not change the thread pitch, to give you the clearance. Even then you would be distorting the thread angles.
My original tutorial was mainly to show the steps, and not really intended to be used for printing. You can use larger numbers of arc segments to smooth out the threads, but you could also cause yourself a world of pain chasing down holes in the model once you do the subtracts. Not having done that myself, this is just an assumption…it may work out perfectly fine. Mechanically, I wouldn’t put any faith in a printed bolt due to delamination once under any sort of tensile force.
Lastly, from experience I have to model the object with a 0.2mm offset to most things I print to get them accurate to the intended size. This is due to shrinkage once the model cools, and also squish as each layer is put down. This can only be determined with trial and error…and each one is different depending on infill density, wall thickness, overall size…etc, etc. It’s never an exact science.
Going by your last comment, you have changed your approach. It would be good to see what your final solution was/is.