One of our nice aluminum water bottles hit the floor and, of course, the tiny little hinge shattered. It’s some wonderful engineering plastic, but just look at the leverage you can apply to those few millimeters of material. This is the sort of repair that can’t possibly be economically justified, but it pisses me right off when something that should be rugged turns out to be this fragile.
The 2 mm steel hinge pin snapped the molded plastic center post of the hinge off the cap; we found the larger fragment, but the smaller one may lurk under the refrigerator for quite some time. Nothing bonds to this plastic and, if the post broke in the first place, adhesive isn’t going to help.
Some doodling showed that a replacement hinge post should be machineable. The general idea was to square up the remaining chunk of the post, then attach a replacement hinge pivot with a screw. The post is almost exactly 1/4-inch thick, call it 6.2 mm, which means the right-angle feature under the pivot ought to keep the whole affair from twisting.
I planned to leave the left side unmachined and cut it to fit by hand, but then figured, eh, just make it happen. I also expected to leave the area around the screw a lot thicker, with a neat counterbore around the head.
This being a bash-to-fit, file-to-hide kind of project, I wrote a snippet of G-Code (at the bottom of the post) to chew out the part from a sheet of Lexan, then did the perpendicular hole & countersinking with manual CNC.
No pix of that; I was working in a white-hot fury. Basically, I double-sticky-taped a slab of Lexan to a sacrificial sheet, clamped it to the tooling plate, and had at it with a 2 mm end mill. Cutting a 6.4 mm sheet with a 2 mm end mill is a bit iffy, as the flutes are just barely that long; the mill was armpit-deep in swarf and I was dribbling water into the cut to keep it cool.
By the time I stopped for a picture, the situation looked like this.
For what it’s worth, that’s the second part. I had to lower the screw head below the top of the half-round feature on the left end in order to clear the cap. That’s what CNC is really good for in my shop: make another one, just like the other one, only different exactly like that.
I drilled a #50 (2-56 tap) hole in the cap pretty much by eye, using laser targeting to touch off.
The hole wound up minutely too far inboard, but some filing cleaned up the stub edge and it was all good. I started the tap in the mill, held loosely in the chuck and turning it with my fingers, then finished up on the bench.
The screw hole goes all the way through the cap. I filed the screw down so the end sits flush at the bottom of the cap, where the silicone rubber gasket should seal firmly against it.
Here’s what the hinge looks like with all the bits assembled. The spring bears on the screw head, which makes the cap open with more snap than before. I put a little counterbore under the screw head, even after lowering it, to reduce the spring tension.
The cap has a spring-loaded latch that never worked very well in the first place and this repair didn’t improve it. As nearly as I can tell, the molded ledge on the cap has a rounded edge that the latch simply cannot engage. This is beyond even my level of interest; Mary was accustomed to using the wire snap to hold the cap closed and that practice will continue.
Works well enough for us and I got some Quality Shop Time on a rainy afternoon.
The G-Code uses a slightly modified & simplified version of the tool length probe routines. I’m not convinced that using the G59.3 coordinate system is the right way to go, but everything else seems worse.
(Water bottle hinge repair) (Ed Nisley - KE4ZNU - June 2010) (Rough-cut 1/4-inch plate with clamp at +Y) (Sacrificial plate below, double-stick tape to secure) (Tool change @ G30 position above length probe) (-- Global dimensions & locations) #<_Stock_Thick> = 6.5 (overall thickness) #<_Traverse_Z> = 1.0 #<_Safe_Z> = 30.0 (clamp clearance) (-- Section controls) #<_Do_Outline> = 1 #<_Do_Drill> = 1 (-------------------) (-- Initialize new tool length at probe switch) ( Assumes G59.3 is still in machine units, returns in G54) #<_Probe_Speed> = 250 (set for something sensible in mm or inch) #<_Probe_Retract> = 1 (ditto) O<Probe_Tool> SUB G49 (clear tool length compensation) G30 (to probe switch) G59.3 (coord system 9) G38.2 Z0 F#<_Probe_Speed> (trip switch on the way down) G91 G0 Z#<_Probe_Retract> (back off the switch) G90 G38.2 Z0 F[#<_Probe_Speed> / 10] (trip switch slowly) #<_ToolZ> = #5063 (save new tool length) G43.1 Z[#<_ToolZ> - #<_ToolRefZ>] (set new length) G54 G30 (return to safe level) O<Probe_Tool> ENDSUB (-------------------) (-- Initialize first tool length at probe switch) O<Probe_Init> SUB #<_ToolRefZ> = 0.0 (set up for first call) O<Probe_Tool> CALL #<_ToolRefZ> = #5063 (save trip point) G43.1 Z0 (tool entered at Z=0, so set it there) O<Probe_Init> ENDSUB (-------------------) (-- Get started ...) G40 G49 G54 G80 G90 G92.1 G94 G97 G98 (reset many things) M5 (msg,Verify clamp to +Y, stock taped down) M0 (msg,Verify X=0 at left edge, Y=0 on finished centerline) M0 (msg,Verify tool touched off at Z=0 on surface) M0 O<Probe_Init> CALL T0 M6 (ensure first tool change pauses) (-- Drill the hinge pin hole) #<Pin_X> = 7.0 #<Pin_Y> = 0.0 #<Drill_Dia> = 2.06 (Drill diameter) #<Drill_Num> = 46 (Drill number) #<Tool_Num> = 146 (Tool number) #<Drill_Radius> = [#<Drill_Dia> / 2] #<Drill_RPM> = 3000 #<Drill_Feed> = [#<Drill_Dia> * 100] #<Drill_Depth> = [#<_Stock_Thick> + 2 * #<Drill_Dia>] O<Doing_Drill> IF [#<_Do_Drill>] (debug,Insert Num #<Drill_Num> drill) T#<Tool_Num> M6 O<Probe_Tool> CALL (debug,Set spindle to #<Drill_RPM>) M0 F#<Drill_Feed> G0 Z#<_Traverse_Z> G83 X#<Pin_X> Y#<Pin_Y> Z[0 - #<Drill_Depth>] R#<_Traverse_Z> Q[2 * #<Drill_Dia>] O<Doing_Drill> ENDIF (-- Mill outline) #<Hinge_Radius> = 3.75 (half-width of hinge body) #<Cutout_Base> = 2.75 #<Cutout_Screw> = 1.50 #<Cutout_Screw_Y> = [#<Hinge_Radius> - #<Cutout_Screw>] #<Cutout_Screw_A> = ASIN [#<Cutout_Screw_Y> / #<Hinge_Radius>] #<Cutout_Screw_X> = [#<Hinge_Radius> * COS [#<Cutout_Screw_A>]] #<Passes> = 3 #<Mill_Dia> = 1.98 (end mill diameter) #<Tool_Num> = 20 #<Mill_Radius> = [#<Mill_Dia> / 2] #<Mill_RPM> = 3000 #<Mill_Feed> = 100 #<Entry_XL> = [0 - #<Mill_Dia>] #<Entry_YL> = [0 - 2 * #<Hinge_Radius>] O<Doing_Outline> IF [#<_Do_Outline>] (debug,Insert #<Mill_Dia> mm end mill) T#<Tool_Num> M6 O<Probe_Tool> CALL (debug,Set spindle to #<Mill_RPM>) M0 F#<Mill_Feed> G0 X0 Y[0 - 2 * #<Hinge_Radius>] (get to comp entry point) G0 Z#<_Traverse_Z> G42.1 D#<Mill_Dia> (cutter comp right) G1 X#<Pin_X> Y[0 - #<Hinge_Radius>] #<Step_Z> = [#<_Stock_Thick> / #<Passes>] #<Current_Z> = [0 - #<Step_Z>] O<Outline_Passes> REPEAT [#<Passes>] G2 J[0 - #<Hinge_Radius>] Z#<Current_Z> (ramp down to cutting level) G3 Y#<Hinge_Radius> J#<Hinge_Radius> G3 X[#<Pin_X> - #<Cutout_Screw_X>] Y#<Cutout_Screw_Y> J[0 - #<Hinge_Radius>] G1 X0 G1 Y[0 - [#<Hinge_Radius> - #<Cutout_Base>]] G1 X#<Pin_X> G1 Y[0 - #<Hinge_Radius>] #<Current_Z> = [#<Current_Z> - #<Step_Z>] O<Outline_Passes> ENDREPEAT G0 Z#<_Safe_Z> G40 O<Doing_Outline> ENDIF G30 (back to tool change position) (msg,Done!) M2
2 thoughts on “Water Bottle Spring Cap Repair”
That’s an awesome fix.
The more I read your fixes list — and the more it inspires me to do the same — the more I feel like there should be a branch of art, probably performance art, showcasing repairs that are economically unjustifiable but philosophically and practically superb.
Last week one of my big box fans had the fan speed knob break. It was a free fan: I found it at a garage sale at the “please just take it away” point in the sale. But it was useful, and it’s a big fan, comparable in size to my entire garbage can. Infuriating to throw away something that big and that useful. Ten minutes of work with the lathe and it has a new knob, that is worth, well, infinitely more than the value of a free, broken fan, but now it’s useful again and I don’t have to throw it away. Before reading your blog I probably would not have considered doing this, but now I take a perverse pride in it, and many of my friends are likewise suddenly thinking “but why should I throw this away?” when things break. (Of course, the downside is they’re asking ME to help, but that’s okay.)
showcasing repairs that are economically unjustifiable but philosophically and practically superb
Sometimes I think I should rename this blog “Mostly, He Fixes Stuff”.
Sort of like the whole Maker movement, except having practical results.
now I take a perverse pride in it
I admit to the same feeling: “I am not going to buy another one just because a ten-cent part broke! Watch this!”
Thanks for the kind words… and spread the meme!
Comments are closed.