This is not really checked in any technical way, just a visual "will it work" type of check. We have had ours questioned before and s now we just put something in there that is meaty enough to look the business.
Having tried to cut down fabrication time by limiting on the fly designing, this part was modeled on computer along with the rest of the car in solidworks. Once we started to fabricate it up we noticed small problems still begin to pop up.
The design is basically a piece of 3mm thick steel plate that has been bent and cut to have an open trapezoidal shape, in which the head rest portion of the sit will sit an the most reward face of the seat should sit flush on the support.
First problem was being able to accurately make the bends. To get the desired radius the idea was to hit the plate around a tube of adequate diameter. Couldn't find a piece the right size so went slightly oversize to the closest I could find.
Next problem was insuring the bent was in the right location, I measured from the model and marked on the plate the end point of the bend arc and attempted to line this up in the vice. I used a spirit level to try and get the part and the temporary bending die at 90 degress to each other.
So the part seemed all set, and bending with the biggest hammer we had was quite quick and didn't take much force. I used an angle gauge to check the part as I bent it after every few strikes with the hammer, again attempting to keep a level datum I used the spirit level across the top of the bending die. It worked ok for what I thought the required accuracey was at this point, really it's just an eye ball check plus or minus 2 to 3 degrees, seemed ok to me.
Then we test fit the part in the car with the race seat in. At this point the race seat is still not mounted correctly because I would need to be able to use the chassis to fibreglass around the bars and we can't take the chassis anyway from the rest of the team for a whole day at this point so we're trying to do the best we can (this is dangerous because you basically can't get the seat to sit right at all unless you're just plain lucky and if you do get it in the right spot you definately won't get it in the right spot twice)
So we take a guess, trying to line the part up using a make shift plumbob to get the centre point, again not very accurate but we figured it was good enough for this particular part. Line it up on the bars and make a mark with a paint pen to show where we need to cut the excess length back for it to fit flush with the roll hoop.
Mistakes all over the place on this stage, not really centrlly placed and to fit it in there without it being cut means shifting the seat slightly out of place to be able to get the plate in there, this changes where the part fits also so is basically a total waste of time as I wold later find out.
Also a mistake made here due to rushing was that the plate was lined up on the front face of the roll hoop bar and not the rear face. The design in solidworks shows the part being almost a flush tagent fit with the rear face, a really basic error made there that does change the fit of the part quite drastically.
The cuts were made and the test fit then immediately showed a major error, the seat sat quite noticably out of position to get it to fit and then once the part was in the correct place there was a very large gap. Some adjustements to the bend angles were made by hand and then some other slight bends at the ends of the part were exerimented with to attempt a retro fit to use what we had now incorrectly fabricated.
This didn't work and at this point the part was looking like a total throw away and start again, it has fundemental failures in it and it will not mount correctly to the seat or the frame and if it is positioned to fit it then moves the seat so far out of position it risks making the drivers seating position not rules compliant.
Another leason learned again about trying to complete work in a rush. The model is so well done that it should have bee trusted and the length of the part made to match it as accurate as we could physically make it. If this was done it would have been picked up that the part was being mounted on the wrong face of the bar and perhaps the subsequent mistakes could have been avoided.
Really not an accpectable turn of events and I wish that it had ot happened, we are planning to salvage what we can of the part but it may be a full re-fabricaton job, there are some attempts being made now to solve the problem but it was a large motivation kiler for the morning.
A few points to myself for next time
- Rule number 1: check the design before positioning
- Rule number 2: check the design before you cut
- Rule number 3: trust the design
- Rule number 4: check the positioning of everything around the part that ineracts with it
- Rule number 5: cut it, make it fit, if it doesn't fit, check the design, if it still doesn't fit.....make it fit.
No comments:
Post a Comment