Screws into deck pulled out...

Sep 8, 2014
2,551
Catalina 22 Swing Keel San Diego
What psi (pounds per square inch) have to do with this?
______________________
800 Psi is the rated strength of threads cut into epoxy resin for strength, as tested by WEST Systems and published in the articles I posted.


I don't mean to quibble. I think the observation that facts by themselves aren't always applicable or useful in understanding what's needed.

The cited published fact means little per se. I respect West Systems, use their stuff, trust them, etc., etc. HOWEVER:
  • Is the cited psi capacity in tension, compression, or shear?
  • Does it mean that you take the projected area in square inches of the threads perpendicular to the long axis of the screw in one direction and measure it against "withdrawal" with an imposed load of x pounds?
  • How is the load applied? Slowly (like if you take the Superglue guy who is hanging from a helmet which has been attached to a square [inch] which he is gently hanging from) or like you snapped it with a sudden load?
  • Is the load applied torsional or just axial?
  • The factoid is only relative; and then maybe relevant or not (i.e. proportional with some safety factor)?
How some fitting gets loaded-up and what the substrate is has a lot to do with whether the materials fail. Examples: mortar and bricks. Work great in compression -- work for shit in tension -- work okay in shear. (if you allow me to generalize.)

Another example, more relevant: In the case of the dodger fitting let's talk about how the fitting is "challenged" and has to respond to forces. In the axis of the two screws that hold the fitting that the dodger is attached to things happen very differently if the load is limited to being in-line with the axis of the screws and the "hinge-pin" that fitting has with the dodger frame. Take the same fitting and bend the dodger framer from the side -- suddenly the "hinge-pin" ceases to just a point load and it becomes one where the frame of the dodger is acting to create a twisting, torsional element that is imparting very different forces. So an 100# load (for example) on the dodger isn't transferring 100# -- but the dodger frame becomes a lever that is twisting and bending the little screws in a very different way.

The "speed" (actually acceleration or deceleration) of the load on the fitting is material. Example, for illustrative purposes, you can take 200 pounds and slowly load up the fitting -- but take 200 pounds and slam it onto the fitting and it just is completely different. Force exerted (or received) is a function of load and acceleration (i.e. time, rate of change of speed, etc.)

If the loading is in a predictable directions and is not subject to dynamic changes, then deducing what is adequate is easier. (e.g., a picture frame hanger which is restraining a load very close to being only in shear will hold a __# item). But approximating what a person being thrown around by a pitching boat or tripping and catching themselves from falling forward, back, or sideways is a much different thing. ;^)))

A seemingly simple dodger which (for example purposes) is strapped back and tightened down and fully secured is a different assembly than one that has it front wind rolled-up, it side curtains not in, or it's retraining lines not taught and running aft to make the sides, front, and top function light a stretched membrane.

Anyway, I've OVER DISCUSSED THIS ISSUE WAYYY TOO MUCH. ;^))) I apologize. Hopefully JViss will have turned his thread alert off.
I can only assume WEST's test results are in Tension, not shear or compression. You didn't over discuss, you make really good points. The area of repair in question is where the Bimini Hinge screws down. Before he posted the pic I had thought it was just the leading edge or side edge of the fabric screwed through grommets. The more I think about that assumption the more silly is sounds... I've seen various attachment methods for Bimini fabric to the deck, usually the turn-key fabric snaps, not screws... anyway, the hinge may experience loads in tension & shear for sure. Regardless, a cap screw set into 2 inches of glass/epoxy threads should be more than adequate.
 
  • Like
Likes: Will Gilmore
Sep 8, 2014
2,551
Catalina 22 Swing Keel San Diego
So a screw in epoxy can take an 800 "PSI" load??? Sounds great, let's see what we get when we figure it out on the back of a napkin. A 1/4-20 screw has an area of (0.25)^2 x 3.1415/4 = 0.049 square inches. What is really happening here is the square inches will be the area of the threads that are in the epoxy since the epoxy will most likely fail before the cross section of the screw breaks. For sake of argument let's assume 6 threads are fully engaged in the epoxy. From this chart http://www.carbidedepot.com/formulas-tap-standard.htm we see the minor diameter is 0.1887 so if we assume for simplicity the threads don't have clearance at the tip or the root the area of one thread is the area of the major diameter minus the area of the minor diameter. The latter is (0.1887)^2 x 3.1415/4 = 0.028 square inches.

So one thread is 0.049 - 0.028 = 0.021 square inches.

Going back to my assumption of 6 threads full engaged we get 0.021 x 6 = 0.126 square inches of thread in the epoxy. 800 PSI x 0.126 SI = 100.8 pounds in tension for the epoxy to fail. Given two screws it (the fitting) might hold up to 200 pounds in tension.

Assuming someone grabbing the dodger to prevent a fall or falling into it will likely apply a severe bending moment to the fitting. A 100 pound person will grossly overload the fitting in this case. The medical examiner and forensic scientists should be able to figure out what went wrong during the accident investigation.

Any of you mechanical engineers or professors feel free to critique my simplistic analysis.
I am not an engineer but there is a flaw in your calculations. The surface area of a 1/4" 20 cap screw cannot be determined by the simple surface area of a cylinder, the threads greatly increase that surface area. Also, you mentioned only 6 threads; take a look at the pic the OP posted... That area is a minimum of two inches deep. I get the feeling some people don't have much faith in this method, but seriously... I'm telling you from experience that there is an insane amount of bite and holding power this way, even for small screws or bolts.
 

SG

.
Feb 11, 2017
1,670
J/Boat J/160 Annapolis
Allowishish: Is the boat outside or inside now? Don't try to fix it if the temps are not within the range for the epoxy to set!!!

If you want any advice, it would be useful to see some photos of how the dodger is rigged: i) when it's Up with the lines, snaps, and fabric all tensioned; and ii) if it's folded down ; and, if it's setup for partial deployment. If you can send some shots of the connections that failed; and, the connections on the side that didn't. Are they both distressed?

There's no rush -- the ice will not be out for another five-or six months :^)))).
 
  • Like
Likes: Will Gilmore
Oct 30, 2017
183
Catalina c 27 Lake Pueblo
It will be on the water all winter.

I am heading down this weekend to see just how badly I screwed it up. But honestly at this point I am trying to polish the proverbial turd. My focus this weekend will be fine fitting the plywood decking for the larger berth and cleaning up the dodger mess.
If it pulls out again I will at that point fully clean it up and do it properly.

That said it is supposed to be low to mid 60's with decent wind so I am hoping to get out and get the sails up as well.