Previously, I’d replaced some bearings in another Duet.
So, this time around, I grabbed a Whirlpool Duet GHW9400PW0 (used, with a known broken spider), and thought I’d give repairing it a shot. As a side-note, purchasing all these non-working bargains has resulted in me having too many working machines.
Note that before you continue reading, it’s worth mentioning that in my case, 2 of the spider arms had cracked off near the center of the spider (3rd arm still connected to the center & shaft). If yours completely disintegrated, you’re in a different situation, and this isn’t going to help you very much (alignment would be a massive issue).
In any case, this certainly isn’t the first GHW9400 out there with a busted spider. A little searching will show it’s a common problem. Whirlpool makes the spiders out of aluminum – not the good kind, but the terribad “pot metal” (or so I’ve read) kind.
Whirlpool doesn’t sell the spiders separately. Not only are they one of the most vulnerable components, but to get a replacement, they force you to buy the entire stainless steel basket at around $400-500. You probably won’t spend that much (and will just buy a new machine for a little bit more). Your old washer goes to the dump. Of course if you break down and buy the whole basket, your old basket goes to the dump. So no matter what you do, something is going to the dump. Internally, I wonder if their motto is “Whirlpool – Killing the environment one machine at a time“.
Anyway. On to the repair.
Disclaimer: This is probably dangerous. I may have done something wrong. Or I may have neglected to mention something vital (I’m writing this from memory). If following anything I say here results in damage, injury, death, or the end of all mankind, I am not taking responsibility. You’re doing this at your own risk. It’s worked for me so far, but it was a “fingers crossed all the way” thing, and for all I know the thing could blow up on you.
Here was the spider’s condition (sorry, didn’t take images – hopefully you’re looking at yours and can see what I’m talking about):
-arm #1 was still “intact” and attached to the shaft/spindle and “center” section
-arm #2 had cracked off from the center
-arm #3 had cracked off from the center
-the basket itself had rubbed through the plastic tub, but was itself, fine
Essentially, everything was still “there” as in “attached to the basket”, hadn’t completely disintegrated, and was more-or-less in position. Obviously when the machine was run, having arm #1 as the only one connecting the spindle to the basket meant it was flexing and wobbling like there was no tomorrow.
NOTE THAT THESE STEPS BEGIN WITH THE SPIDER STILL CONNECTED TO THE STAINLESS BASIN
STEP 0 – CLEANING & DRYING
The thing was covered in white gunk (aluminum’s version of rust), soap scum, and much else. Used high pressure water, “Simple Green”, sandpaper, and scrubbing to clean it up.
Rinsed everything and dried under a heater.
STEP 1 – putting everything into position (risky):
Why is this risky? Assuming you’re in the same situation and looking at your spider/basket, if you grip the shaft, you can probably pull it towards you, or push it away. The remaining arm that’s still connected to the shaft (arm #1) flexes. There’s no method I know of to figure out how pushed in, or pulled out it actually should be.
In my case, based on the damage to the *tub* during operation (concentrated at the front of the machine), I surmised that the arm had pulled/flexed away from the basket and was probably sitting too far away from the basket at rest.
I pushed it towards the tub, and did so with the other arms. Fortunately, cracks tend to be zig-zaggy, and the pieces fit together. They ended up lining up pretty well at the crack.
To hold them all down at an even level (yet pushed down), I threw the concrete weights from the tub on 2 of the arms.
STEP 2 – measurements/alignment
I had to make sure that the spindle wasn’t cockeyed. Obviously, if it were, the basket would rotate all wonky once everything was done (and inevitably break + smash things). I took a tape measure. Here’s a paint image to give you an idea:
(Remember that the spider itself is still connected to the stainless steel basket. Obviously this paint image doesn’t show the cracked arms, but if you mentally “snap off” 2 of the arms, you’ll get an idea as to where the cracks are.)
The yellow/orange lines are where I used the tape measure. I measured the distance from the nut (which I loosely threaded on the top) to the edges of all 3 spider arms. DISTANCE MUST BE THE SAME FOR EACH – if it’s not, the spindle is almost certainly cockeyed and bad things are pretty much guaranteed to happen.
All distances were the same.
STEP 3 – PRE-MENDING THE CRACKED OFF / BROKEN ARMS
Ideally, if you’re doing this, you should probably use a welder. We have one, but I didn’t use it because (1) i’m successful when welding steel, but i usually end up simply melting new holes in aluminum. (2) I didn’t have the aluminum wire for the mig (and the gas you need).
I used a combination of JB Weld and Fibreglass.
I started with JB Weld. If you decide to go my route, grab the “real” stuff (red/black tubes), and grab a lot. Don’t use the JB “Quick” Weld, as it’s not as strong, and you need every advantage you can.
I Mixed the JB Weld, used a toothpick to push it in the cracks, then used my finger to smooth it over.
Then I waited for it to cure.
Next, I cut strips of fibreglass, contoured to follow the “perimeter” facing me. This stuff is a pain to work with, but I figured the JB Weld on it’s own probably wouldn’t hold up forever (if at all), and like I said earlier, you need to take every advantage you can.
Mixed the resin, doused the fibreglass in it, and placed the fibreglass in position, making sure that I crosssed the part of the “perimeter” where the breaks/cracks (and now JB Weld) were. The thinking was that if the fibreglass crossed the cracks and I made use of all the surface area I could, it could only help if the JB Weld wasn’t up-to-the-task.
I let the fibreglass dry and then smothered the entire thing (including over the fibreglass I’d just put on) with more JB Weld.
Beyond giving the JB Weld as much surface area to work with as possible (and smoothing out over the fiberglass), so much of the arm had deteriorated that I figured if by some chance this worked, the arms were likely to break off somewhere else soon anyway. May as well get a pre-emptive jump on things.
So everything I could see/access so far (the other side wasn’t accessible due to the stainless tub) was covered in JB Weld and fiberglass.
I let it all dry a loong while.
STEP 4 – REMOVING THE ARM FROM THE BASKET
Obviously doing 1 side of the spider probably wasn’t going to be enough. Time to remove the spider so I could get to the other side.
This was the first test of strength for the spider, as it has to be pried out. To remove it from the stainless basket, there are a total of 6 screws holding the spider in.
Once those screws are out, you have to pry in 2 places at once – I picked a spider arm and used a screwdriver to pry between the basket lip & the spider arm (because the screw-holes are beveled in, you have to force a gap between the arm & lip), and pried the arm itself up at the same time with a prybar.
This is quite a bit of force being exerted – fortunately, I didn’t hear any cracking. The spider was already pretty solid.
STEP 5 – THE OTHER SIDE
With the spider (now one piece more-or-less) out of the basket, I had access to the opposite side. I began by cleaning it up (washing, cleaning, sanding, scrubbing) and drying it.
You can guess what I did next. That side pretty much took a bath in JB Weld, starting with the cracks, letting it dry, and then smearing JB Weld over the rest. I ignored the far edges because most pictures out there show cracks/breaks being concentrated around the center, and I was already running low on the JB.
More dry/cure time.
STEP 6 – PAINTING THE SPIDER
Since I hadn’t used JB all the way to the edges, I figured paint might help to keep water/soap/minerals from attacking the little aluminum that was exposed (and perhaps catch areas I didn’t quite perfectly cover with the JB.
I painted the entire thing with some pretty standard Tremclad.
…and more dry time.
STEP 7 – REATTACHING THE SPIDER – SILICON CITY
As you’ve probably figured out, throughout the process I had days of drying time to ponder/plan. During this down time, I wondered why the spider was never attached to a more central part of the stainless basket to any degree (why limited to the edges?). Was the arm supposed to flex a little and absorb shock? Was the tub expected to flex? Dissimilar metals issues? Physics/leverage? Oversight?
I almost considered drilling through some of the “compartments” in the spider, right through the stainless basket and bolting the 2 directly together in a few spots for added strength, but because there were so many possible reasons it may intentionally *not* have been that way, I decided to go with a flexible compromise.
I grabbed 3 tubes of the Permatex silicon stuff (1 silicon, 1 RTV high temp gasket stuff, 1 black RTV gasket thing), dumped all of 1 tube in the middle of the basket, and the other tubes in thick lines down the “arm” runs.
Then I plopped the spider back on, sandwiching the silicon down between the two.
Silicon Notion was 3-fold:
- the silicon (when cured) should transfer *some* of the energy directly to the basket, starting at the center, which *should* result in slightly less energy being transferred down the arms.
- if the spider actually shouldn’t be direct-connected to the basket anywhere but the edge, this stuff would/should flex enough not to hurt things too badly.
- in the event this whole endeavor failed and the arms were going to crack off again anyway, the silicon would hopefully at least keep the arms from instantly slamming in the wrong direction – keep them kinda in place so that instead of a big boom, I get a more gradual crunch-crunch-bang.
The biggest concern I had here was that the silicon might not adhere super-well, go flying off at high spin and end up plugging the drain system. Figured it was a relatively small risk for the possible potential.
STEP 8 – REASSEMBLY
After letting the silicon cure (more waiting….), I started reassembling.
Once back in the plastic tub, I rotated the stainless basket by hand to verify it wasn’t cockeyed, and to make sure it wasn’t rubbing the tub (if I had a problem here, I’d have to snap everything back apart and start over).
Fortunately, all seemed well.
I tried prying between the basket + tub, and there was no play, and no cracking.
Things were looking good.
- Assembled the rest of the machine.
- Ran a number of empty test cycles (using a flashlight to watch the space between the stainless + tub the whole time to ensure it wasn’t coming loose).
- Made sure it seemed balanced. Listed for noises, etc.
- Ran cycles with clothes (again, watched with flashlight).
- Checked the drain catch to make sure I didn’t have JB Weld, paint, fibreglass, or silicon in there.
- That was it!
As far as I can tell, it’s worked fine. Of course, it all could end in a catastrophic bang, but I’m optimistic.
One thing I have avoided is the use of the “Extra High Speed” spin – I ran it once, that particular load wasn’t balanced (washer tried a few times to rebalance) and the thing was banging away – I realized I couldn’t tell whether it was the load or if things were coming apart, so I shut down the washer, checked the basket for integrity/play, and when I found it was still solid, I figured I’d rather stick to “High Speed” instead and forgo the worry every time I hear a noise.
I probably should have allowed the JB Weld (everything really) to cure for a week. While the stuff should generally be done within 24 hours, it seems to continue curing & getting harder after that.
If I were to do it again, I’d use real weld the next time and just learn to deal with welding aluminum. Going the JB Weld route seemed like the quicker/easier/cheaper solution at first, but as you’ve seen, I spent days waiting for stuff to dry, and resorted to fibreglass + silicon to help ensure integrity. Maybe I didn’t need the fiberglass/silicon, but to my knowledge, nobody’s tried to repair one of these things before and I had to operate under the assumption that whatever I was doing might-not-be-quite-enough.
In hindsight it would have been much better to weld the thing, maybe use JB to fill out any holes/gaps & weak areas, and then just paint it and go. Welding is quick, and when done properly, incredibly strong – much stronger than any epoxy ever has been in my experience anyway.
In any case, if your spider is gone and you’re looking to repair it, it may certainly be possible, though whether it’s worth the time & effort is questionable. Of course, longevity’s still a question too. As you can probably tell, I’d be inclined to suggest a proper weld rather than the precarious JB/Fibreglass method (still no guarantees, but should be easier, quicker, & more reliable), but I realize that starts to become a “do as I say, not as I do” thing.
Regardless, if you decide to tackle a repair on your spider, hopefully something above has helped you.