Guide rail saw tracking when connecting rails

[nclemmons]

Two thoughts:

1  - what is a reasonable 'tolerance' for rip cuts for woodworking with sheet goods?  In the case of DrD's measurements, .006 is about 2 sheets of paper thick.  For case goods made with plywood or MDF and then sanded, planed, joined and put into an installation, what should we expect in tolerance for our work?  I have and use a 6" digital caliper when running my planer and thicknessing.  Not so much for sheet goods, however.

I have done some projects like my chessboard in the projects area that I hand planed to .001 tolerance.  I've done metal lathe or milling machine work with a tolerance of .001 or higher on occasion.  But on most woodwork, I don't get that close.  And still strive for 'museum quality' in my work.  Frankly, the finger can feel more of a difference than the eye can see sometimes.

2 - what should we expect for guide rails in terms of tolerance since they are extrusions through a die that may vary over time / with wear, etc.  I wonder what Festool considers as agreeable tolerances?  I found a book on extrusions and tolerances - http://www.aec.org/?page=lib_designmanual - but wonder what Festool considers acceptable.

 

[Klemm]

extrusions will vary dimensionally. Having spent more than a few days and nights in extrusion plants (US and China) I can tell you it is a moving target - unlike CNC machining.
In addition to die wear you also have the effect of the drawbench tensioning (stretching) the raw extruded shape. Last, but not least, heat treating can put a spin on dimensions (flatness and straightness).

The solution: if something does not line up - get other extrusions until you have a fit - that's real life. The call for perfection in extrusion tolerances does not promise to be very productive.

We can't wait for FESTOOL to improve their tolerances, so we will change extrusions until we get a match.
Hans

 

[Cheese]

Went to the shop and took some measurements; this was no where close to a gauge R&R for those of you familiar with such.  Rather I took careful single (non-replicated) measurements at each end of my rails and in the middle.  Results:

  Rail      Purchase Date    Purchase Location    End with Lettering    Middle    Opposite End  - all measurements in inches
FS800          08/20/2013          Chicago, IL            0.634                0.635      0.633
FS/1080      04/13/2014          Bob Marino            0.636                0.638      0.636
1400/LR32  03/06/2013          Lombard, IL            0.633                0.634      0.632
FS1400        04/12/2013          Bob Marino            0.636                0.635      0.635
FS1900        05/27/2014          Bob Marino            0.635                0.634      0.632

Those are really, really tight groupings. I'm actually very surprised they group that closely.
FS 800...range = +/-.001"
FS 1080...range = +/-.001"
1400/LR 32...+/-.001"
FS 1400...+/-.0005"
FS 1900...+/-.0015"

Considering that different extrusion dies have been used over the years, and they have probably been in various states of condition, that new extrusion dies have probably been introduced to manufacturing because of tolerance issues, and the rails have probably been manufactured in different manufacturing facilities, subject to slightly different aluminum chemistries, and also subject to slightly different QA techniques, and probably slightly different heat treating/tempering conditions, I find the above results pretty damn impressive. And as [member=19075]DrD[/member] stated, he's had no issue with joining the above rails.

I'd be more interested in finding out what the dimensions of the 2 rails are in regards to the rails that [member=59578]CADru[/member] originally tried to join together. That would provide more information that may resolve this issue

 

[Svar]

It would not be difficult to machine one side of the guide rib after extruding but before anodizing.

 

[DrD]

[member=44099]Cheese[/member]

Getting out on really thin ice, I did a quick analysis (tTest) and found absolutely no difference within a rail nor among all rails.  Now, this must not be considered as a statistical absolute, given the absurdly small sample size with no replications.  However, when considered with the anecdotal evidence of no difference in saw behavior on any of the rails, singly or connected, an inference can be made that the measurements are solid, and this collection of rails are all the same as regards this one dimensional characteristic.

So now, the question is, as you pointed out, what are the measurements for the op's rails.  If different, then something has changed.

 

[antss]

Sorry guys for giving you the wrong impression and sidetracking the discussion.

If you"ll re read my earlier post :  "Your policy needs to be to exchange rails that don't mate up because YOU couldn't maintain a consistent manufacturing facility"

My point is Festool needs to make good on (read: exchange) rails that don't match up.  Not to have a band aid suggested by the trainers, no matter how capable.

Seemingly , Brice, Svar, Hans and even DrD are in agreement with me on this point. 

 

[DrD]

[member=59331]Hans Friedebach[/member]

Don't forget that radiator fins are largely extruded aluminum and there are very tight tolerances with those regarding pore/hole size, fin thickness, etc, and these tolerances must be held not only for the entire billet being extruded, but also among billets, over continuous production. 

While working with aluminum can be picky, it is routinely done to quite amazing tolerances and reproducibility - far beyond what would be needed to make guide rails work together for saw or router use.

DrD

 

[DrD]

[member=727]antss[/member]

"... and even DrD..."  I didn't know our agreement on issue(s) was uncommon.  It's always good to have thorough discussion and evaluation of any issue/situation.  And, I do agree the op should get a replacement rail - he might want to provide Festool with the channel measurements of his rails, to ensure the replacement performs satisfactorily.

And, I do agree that if his measurements are significantly different from rail to rail, and from measurements submitted by other FOGgers (including mine) there must be root cause analysis, and permanent corrective action taken by Festool.

Are we still in agreement? [smile]

 

[DrD]

[member=59578]CADru[/member]

Can/will you take the top channel width measurement at each end and in the center of your rails and post them?

Thanks,

DrD

 

[Cheese]

However, when considered with the anecdotal evidence of no difference in saw behavior on any of the rails, singly or connected, an inference can be made that the measurements are solid, and this collection of rails are all the same as regards this one dimensional characteristic.

So now, the question is, as you pointed out, what are the measurements for the op's rails.  If different, then something has changed.

FWIW...I took a quick measurement of the 3 rails that I own, purchased over 4 years from various suppliers, and found them all to be within the .632 to .635 range. Again, another very tight grouping.

I'm trying to understand the tolerance variance that renders the rails as being unsuitable for joining.

 

[DrD]

[member=44099]Cheese[/member]

Yeah, you're asking the tough question - when does the rail to rail channel width/thickness mean and variation become noticeable to the saw.  This then will relate to the "tightness" of the gibs on the saw base, pushing the front mating channel on the saw base into the front channel of the rail, and the saw bottom plate being pulled into the rear face of the rails.  Once set for one rail, any looseness or increased tightness when the saw enters the second rail would seem to indicate the second rail's mating channel was not nominally the same width/thickness as that of the rail to which the saw was adjusted.

Also, might dimensional variations in the underneath channels of the 2 rails effectively skew the alignment of the top channels to each other if the bottom channels are connected first?  If so, does this suggest the possibility of placement of the connectors and sequence and methods by which they are secured being contributors to the issue?  Of course, this all becomes moot if ALL rails are extruded to spec (whatever that may be).

It's late, I'm going back to bed; I'll see what's here tomorrow.

Don

 

[DrD]

[member=44099]Cheese[/member]

One more thing what brand of café bike is that in your avatar?  I drive a Honda VT1100C bagger.

 

[Cheese]

It's late, I'm going back to bed; I'll see what's here tomorrow.

Don, not much here today...seems like we bored the others out of their skulls last night. [eek]

That was originally a 1973 Norton Interstate that I purchased new, drove it in that configuration for a year and then changed it over to a Commando variant and then decided to cafe it. The cool thing is that the cafe parts are all circa 1974-1975. Rear-sets, bars, Dunstall exhaust, Borrani alum wheels, fairing, Megacycle cam, heads, pinstriped paint and gold-leafing, yada, yada are all old school.

I've got the oil tank off of her now because I'm chasing down...wait for it...an oil leak. [doh] [jawdrop] [doh]

 

[DrD]

[member=44099]Cheese[/member]

Yeah, I recon we did.  But you know what, it beat just whining about the issue.  Every now and then there must be a technical approach to problem causation and solving, not just lip service.

Now about you're café racer - wow, wow, wow; I may even be totally jealous.  Not really jealous, I'm just glad you've got such a FINE ride - bet it sounds awesome!

Well if others see the rail problem or possibly the underlying problem of product quality as an issue(s), I'll step back and let them take wherever.  I am grateful for your  input; especially good to see your numbers correlated with mine.  Guess I spent too much time getting paid to do this and being successful in doing this back in the world.

 

[Brice Burrell]

It's late, I'm going back to bed; I'll see what's here tomorrow.

Don, not much here today...seems like we bored the others out of their skulls last night. [eek] .......

I wasn't borded out of my skull until you guys started talking about you bikes. [tongue]

All joking aside, I don't care much as to the why the rails are out of spec.  That's Festool's problem to sort out, and they need to sort it out.  Very high prices tags come with very high expectations. 

BTW, my LR32 rails don't mate up well and it is a total PIA drilling holes.  As the first gib on the LR32 base hits the wider rail it is much harder to push along.  So much so you think it is locked into one of the holes causing you to misdrill a hole if you aren't extremely vigilant.

 

[Cheese]

Very high prices tags come with very high expectations. 

BTW, my LR32 rails don't mate up well and it is a total PIA drilling holes.  As the first gib on the LR32 base hits the wider rail it is much harder to push along.  So much so you think it is locked into one of the holes causing you to misdrill a hole if you aren't extremely vigilant.

Very high prices tags come with very high expectations.   So very true... [thumbs up]

Interestingly enough, of the 2 rails that I connect together all of the time, one is a standard 1400 and the other is a 1400 LR32. I've had no problems connecting or using this combo. You can feel a slight difference in saw movement (3 different saws) as they travel
down the rails, but that's it. 

 

[DrD]

[member=1146]Brice Burrell[/member]

Brice, great to have your input - just finished reviewing your supplemental manual on Kapex, a great resource, Thank You.

So, do you just tough through it when using your LR32 with connected rails?  Does that throw off your hole spacing distance from the reference edge?

Always welcome your insights - even if you're not a bike driver [big grin]

Don

 

[CADru]

[member=59578]CADru[/member]

Can/will you take the top channel width measurement at each end and in the center of your rails and post them?

Thanks,

DrD

[member=19075]DrD[/member]

Here are my rail measurements. The first 2 rails in the list are the rails mentioned in the OP. All my rails were puchased from same local dealer within a 2-3 month time frame, Q1 2016. I don't have a purchase record for the 1400 listed first below. This rail was sent to me when I sent saw in for repair. This 1400 is ~0.008 smaller then the other rails.

Rail: End w/Lettering - Middle - Opposite End
1400: 0.627 - 0.626 - 0.627
1400/LR32: 0.633 - 0.634 - 0.634
Additional rails
800: 0.634 - 0.634 - 0.634
1080: 0.634 - 0.634 - 0.634
1900: 0.634 - 0.633 - 0.634
2700: 0.633 - 0.634 - 0.633

 

[CADru]

Don, not much here today...seems like we bored the others out of their skulls last night. [eek]

[member=44099]Cheese[/member]
Boring tool talk... never heard of such a thing  [big grin]

 

[Michael Kellough]

The tracking spine is supposed to be 16mm. I'd measure using mm rather than inches.

I have a bunch of the older guide rails (single groove) and they were a pain to join not only because they could use only one connector but also because the width of the tracking spines on the rails varied from 15.8mm to 16.1mm (from one rail to another not from end to end).

It seems it isn't practical to expect Festool to maintain strict tolerance on the rail extrusions but maybe wouldn't be so difficult for them to measure tracking spines so we could order something within 1/10mm of what we want to match. They could make simple go/no-go gauges at 15.9, 16.0 and 16.1mm (and reject anything above or below that range) and label the rails for size.