[Halld-tracking-hw] Carbn Fiber Parts

[Tim Whitlatch]

Hello All,

Now that we have got all the issues out, I propose we have a video 
conference Monday 2pm to discuss our next moves. Does that work for all 
interested parties?
Enjoy the weekend!
Thanks,

       Tim

Curtis A. Meyer wrote:
> Hi Everyone -
>
>     yes, we are all planning to measure the position of the crimp pin 
> after stringing to get
> the best possible accuracy. I am not contesting that. However, making 
> sure that that things
> are as close as possible to where they should be should be our 
> starting point.
>
>     When we were setting all the specifications on the parts and hole, 
> CMU wanted things
> at the 1/2 mil level, but we had to back off due to cost. That said, 
> we had somewhat
> reasonable expectation that we would fall somewhere near the middle of 
> the specs
> with a distribution (like the Al parts) versus what we have with the 
> Carbon fiber where
> we have parts just out side the specifications on the low side.
>
>     When I take a donut and a feedthru, put them through the endplate 
> and feel the substantial
> wiggle that is present. I a concerned to the point where I loose sleep 
> (seriously) over the
> thought of continuing with these parts.
>
>      As per the gas tightness, I believe that our small number of 
> controlled tests on a bench
> appear to have sealed the donut to the feedthrough, but looking into 
> the straw at the glue
> joint, there is some glue oozed into the straw. This is indicative of 
> what we saw when
> we had gas leaks. We can also glue the feedthru to the endplate 
> sample, it is difficult to
> check gas tightness, but we can look. However, even with that done, I 
> am worried about
> guaranteeing all the seals holding in production.
>
>      Our procedure, while controlled very well with the pneumatic 
> gluing and the entry and
> exhaust ports is very good, but if we have wiggle in the parts during 
> installation, we have
> trouble controlling things and I am very concerned about developing 
> gas leaks during the
> experiment. I am not willing to say that we can do it with what we have.
>
>      Tim's comment on the straws fitting is a good one. I also 
> recalled a tighter spacing
> than the 9mil he recalls. That makes this much less of a problem.
>
>      Curtis
> On 7/16/10 2:49 PM, Tim Whitlatch wrote:
>> Hi All,
>>
>> I will state what appears to be the 4 concerns and address each in 
>> order and offer a solution.
>>
>> 1) Uncertainty in the Position of wire.
>>
>> If all allowable tolerances are added up including the location of 
>> end plate holes, the uncertainty of wire position can be as much as 
>> .002+.002+.0025+.001+.003+.0025 = .013 inches.The uncertainty in the 
>> stereo wire positions can be as much as .014 inches. This includes 
>> all maximum clearance between mating parts and end plate hole locations
>>
>> *Possible solution*: As discussed last summer, the plan is to use 
>> either photogrammetry or a portable coordinate measurement machine to 
>> determine the as built locations of each crimp pin. This can probably 
>> get us to within knowing the actual  location within  about.003 inches.
>>
>> 2) Concern with gas seepage at joints.
>>
>> >From the average numbers supplied by CMU this morning;
>> Clearance Fit between donut and straw is within .0027 with pin gage 
>> in place inches (.0032 without)* (drawings allow .004 inches)
>> Clearance Fit between donut and feedthru is .0028 inches (drawings 
>> allow .004)
>> Clearance Fit between feedthru and endplate is .0026 inches (drawings 
>> allow .0025)
>>
>> *Looks like the donut gets squeezed by .5 mil without pin gage while 
>> measuring
>>
>> >From above we note that the actual clearances are equal (within 
>> .0001 inches) or less than what was allowed by the drawings.
>>
>> *Possible solution:* It is important that we test the components in 
>> hand using the practice carbon end plate supplied by Meyer tool. Glue 
>> and leak check. This will tell us whether we have a problem or not.
>>
>> 3) Concern that the last straw will not be able to be installed due 
>> to lack of space.
>>
>> If allowable tolerances are added up for possible straw tube location 
>> discrepancies, we get .002+.002+.0025 = .0065 inches. For stereo 
>> holes this goes up to .0075. The minimum nominal clearance between 
>> any 2 straws is .009 inches (assuming maximum possible straw tube OD 
>> of .622 inches). Therefore, this has already been taken into 
>> consideration and does not pose a problem.
>>
>> *Possible Solution: *The last 3 straws in any row can be fit checked 
>> prior to gluing them in order to avoid any remote possibility that 
>> the last does not fit.
>>
>> 4) Crimp pin fit too tight
>>
>> *Solution:* Have vendor correct this.
>>
>> Cheers,
>>
>>        Tim
>>
>> Curtis A. Meyer wrote:
>>> Hi Beni
>>>
>>>     run-out refers to the build up off all the errors as the parts are 
>>> put together.
>>> Since they are all in the same direction, we get a build up rather than 
>>> a randomization.
>>> I am not referring to glue seeping out, that does not happen.  I'll try 
>>> to restat what we
>>> feel the critical issue for us are that concern us a great deal with the 
>>> parts as they stand.
>>>
>>> The wire position relative to the endplate is indeed controlled by three 
>>> fits:
>>>
>>>     1) Feedthru into the endplate that we know is a sloppy fit right 
>>> now. It was comfortable
>>>         with these parts back in May, but they were shrunk by about 1.5 
>>> mils.
>>>     2) The pin-holder into the feed-thru which is probably ok, although 
>>> is ok because
>>>          the two ridges on the side make it snug.
>>>     3)  The pin into the pin holder, which currently does not fit, so 
>>> this has to be changed.
>>>
>>>      We think that this leads to 3-5 mil (75 to 125 micron) uncertainty 
>>> in the position.
>>>
>>>    The second issue we are concerned about is the gas seal at the 
>>> downstream endplate.
>>> This is where gas is fed into the straws, and if it simply leaks into 
>>> the large volume around
>>> the straws here, we will not be able to guarantee what the gas is in the 
>>> straws themselves.
>>> We know from our pretty serious looks at this with our prototypes that 
>>> the fit of the donut
>>> into the straw and the feedthru into the donut are the major problem 
>>> areas and we had to
>>> have things better than what we have with the current parts. Thus my 
>>> suggestion for the
>>> 1 to 1.5 mil change in size.  We also have a potential leak point where 
>>> the feed through
>>> goes through the endplate. This is a loose fit now. This gluing is done 
>>> by "slathering" glue
>>> on the feedthru, pushing it through the endplate into the donut, 
>>> cleaning up the glue around the
>>> base at the endplate, then injecting glue into the port into the 
>>> feedthru-donut seal. Thus, I am
>>> concerned about the seal from the endplate into the gas volume around 
>>> the straws as well.
>>>
>>>     Finally, the layers with close pack are reasonably tight with the 
>>> straw positions. We left
>>> about 5 mil between straws. Where this hits us is when we go around a 
>>> layer and then try
>>> to fit the "last straw" in. If things have slid due to all the run-outs 
>>> plus the errors in the hole
>>> positions themselves, then it may be impossible to load the last straw 
>>> in some layers without
>>> damaging the chamber. I am not certain what our strategy to deal with 
>>> this would be other than
>>> to plug the holes, don't insert the straw, and try to use the close-pack 
>>> layers to make the straw-to-straw
>>> glue joints.
>>>
>>>      We have been talking about alot since Tuesday and our conclusion 
>>> here is that we feel with these
>>> three factors the risk is bigger than we are willing to accept.
>>>
>>>      Curtis
>>>
>>>    Hi Curtis,
>>> I am a little confused here. What do you mean by run-out? Is this the 
>>> glue that
>>> potentially can seep out? I also do not see where this 5 mil number for 
>>> the wire
>>> location is coming from. I am totally confused. First you glue the donut 
>>> into the
>>> straw and let  it cure.  So having the donuts  undersized by 0.6mil  in 
>>> diamter will lead
>>> to a variation of not more that 0.3 mil in the  location of the center 
>>> of the donut
>>> that defines the location of the wire at this point. Then you put the 
>>> straw into the
>>> frame and glue the feed through into it. At this point an 0.3mil toot 
>>> small feedthru
>>> will cause a shift of 0.15mil in the central positon which defines the 
>>> wire location.
>>> So now I am at max about 0.45 mil  off center for the wire location or 
>>> the crimp
>>> pin holder if you wish.  I do not see how you could get to 5 mil.
>>> Secondly this run-out if I understand correctly that this is glue 
>>> seeping out somewhere.
>>> To try to mitigate such a problem by tighter tolerances is the wrong 
>>> approach. If this
>>> is really a problem then the viscosity of the glue has to be changed. In 
>>> addition you do
>>> not glue in only one straw at a time but many. One question here: When 
>>> you glue in
>>> the straws into the frame is the frame vertically or horizontally. If it 
>>> is vertically this
>>> isue of seeping glue to neighbouring straws is very unlikely. If 
>>> horizontally then start
>>> from the bottom and work up on both sides of the ring then any seeping 
>>> will flow down
>>> and only affect already installed straws.
>>> Before doing any more modification I highly recommend to do some test 
>>> glueing with
>>> straws and the carbon fire spare plate you used to test the reaming. If 
>>> all fits why try
>>> to improve on something that works. We should try to optimize also on 
>>> the cost/benefit
>>> ratio.
>>>
>>> cheers,
>>> Beni
>>>
>>>
>>>   
>
>
> -- 
> Prof. Curtis A. Meyer		Department of Physics
> Phone:	(412) 268-2745		Carnegie Mellon University
> Fax:	(412) 681-0648		Pittsburgh PA 15213-3890 
> cmeyer at ernest.phys.cmu.edu	http://www.curtismeyer.com/      
>
>   
-------------- next part --------------
An HTML attachment was scrubbed...
URL: https://mailman.jlab.org/pipermail/halld-tracking-hw/attachments/20100716/9b7f070e/attachment.html 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: whitey.vcf
Type: text/x-vcard
Size: 131 bytes
Desc: not available
Url : https://mailman.jlab.org/pipermail/halld-tracking-hw/attachments/20100716/9b7f070e/attachment.vcf