Cosman "piston fit drawer", a good idea?

Disagree 100%

Pam Niedermayer - Austin, TX

>Geesh, Paul, I think there's very little cynacism here. The only time anyone complains about someone else making money is when there's a conflict of interest involved.

Pam

Confirming Cosman's technique

David Charlesworth

>I was taught by a friend of Edward Barnsley, and discussed drawer fitting with Alan Peters, at cosiderable length, at the time I wrote the articles in my second book.

Rob is spot on on all counts.

A conventional drawer opening does not change its width or height, with changing humidity.

A solid drawer front does not change it's length but does change in height. As long as the drawer sides are made from quartered stuff, which will not cup, sideways clearance is highly undesirable. Very little is needed to spoil the smooth running and induce racking and sticking. There MUST be sufficient height clearance to cope with the seasonal changes in the drawer's height. Hoadley has a good section on this.

I believe the Karl Malmsten school (where Krenov learned), use a 4 thou" feeler gauge for side clearance, but in my workshop on smaller drawers the clearance is only a shaving or so. There is no reason for this sideways fit to change with time. Most problems of drawer fitting result from poor carcase work, and straightening up the sides makes a huge difference.

I would also agree that very few modern workshops, work to these tolerances, only the very best, and it is wonderful to feel a really precise drawer action.

David Charlesworth

Forgot one thing...

HC Sakman @ Victoria, BC

>"The advantage would be no racking or side to side slop that might cause it to bind/rack" and the drawer front remains flush/parrarel with the carcass/case whenever it's shut.

Chico...

Re: Forgot one thing...

Pam Niedermayer - Austin, TX

>This is more easily accomplished by using center - or in the case of a wide drawer, two "center" - sliding dovetail mounts. Not only do they slide easily for decades and/or centuries, they are smooth, with no racking or other bad things. They're also invisible to the normal user. They also provide the option of making the drawer front inset or not.

Pam

Absolutely Pam...

HC Sakman @ Victoria, BC

>...but that's another style and technique alltogether. As the "cat skinning saying" goes, in a nutshell I can think of more than a dozen different ways to construct and operate a drawer for many different styles.

Basically what you're suggesting is a different way of designing how a drawer operates, but looks the same from outside.

Chico...

You Forgot one thing...

Todd Hughes

>.....I didn't ask for the advantage of a well fitted drawer but I asked for the advantages of an airtight so called "Piston Fitted" drawer

I'm not saying that sloppy ill fitted drawers are good enough....I'm asking are an airtight "piston fitted" drawer that may bind with changes in the weather really needed and are of any real practical advantages. So far I have heard here the advantages are that they are basicaly real cool to push in and out [long as they don't stck I guess] and are a good way to impress people that your are a top shelf woodworker. Wouldn't it be better to build just a well fitted drawer and not worry about making it an airtight "Piston Fitted drawer?....Todd

Why you need no expansion space horizonally...

Paul Kierstead

>For side to side, consider this:

In his construction, a solid wood case, the grain runs the same direction across the face of the drawer and the drawer divider and the top and bottom; also, all are (very nearly so) the same width and the same species. In this case, all will experience very nearly the exact same expansion and contraction so, in fact, zero expansion and contraction space is needed.

However, vertically the sides typically have grain running top to bottom, cross grained to the drawer, hence space is required.

Is interest that if the carcase were made of different species, or -- much more so -- some kind of composite material which did not expand/contract appreciably, it might be a whole different kettle of fish. As they say back home.

Re: Forgot one thing...

Paul Kierstead

>True, they work well, but they rob you of some space, especially significant in smaller work. Additionally, in a few old pieces I have seen with such runners, they had a fair bit of play and racked a little, but it is difficult to say if this was due to very long term shrinkage or original loose tolerances.

And very closely fitted drawers also often (though not always) look quite nice too.

Nonethey less, as Chico says, it is another way to skin the cat.

Re: You Forgot one thing...

Paul Kierstead

>Ok, outside of you, who said they were airtight?

Somehow the numbers don't work

Denis Ch�nard, Orl�ans, Ont.

>GIVEN THAT the drawer opening doesn't change in size

GIVEN THAT wood doesn't move significantly along its length

GIVEN THAT drawer sides are made of quarterwawn stock

THEN the numbers don't add up.

Let's use this for an example:

- take a drawer of a certain width (which doesn't matter for our purpose)

- make the drawer sides out of quartersawn maple (which is commonly used for this application), 3/8" thick

- drawer is dovetailed back and front, as good quality drawers should be

- drawer opening is joined together using sliding dovetails (at least at the front)

If one were to build this drawer and surrounding box in the winter, a piston fit would bind in the summer.

The drawer sides, being quartersawn, would expand sideways (i.e. push out of the box) at the tangential expansion rate. The drawer front and back don't move, and the dovetail joints block the sides from expanding inwards. The box dimentions don't change, the sliding dovetails prevent this. So we are left with only the drawer sides that can change dimensions going from dry winter to muggy summer.

With our example, the drawer sides each would get thicker by 0.00795" going from 6% to 12% moisture content (computed using the WoodNet calculator). Let's round this to 0.008". Add the two sides together and the box should be 0.016" wider in the summer than in the winter. Neither the Malmsten 0.004" feeler gauge test nor a couple shaving thickness can create enough space for the drawer side expansion...

Even if one used more stable mahogany instead of maple, the numbers still wouldn't add. Maybe it works in England because of a less severe moisture gradient over the year, but in most areas of North America I don't believe the kind of clearance you use would work.

I'm not arguing with your results, but either I'm missing something or your level of side clearance will not work over here.

Cheers,

DC

Re: Somehow the numbers don't work

Dan Clermont in Burnaby

>Hi Denis,

The Quartersawn sides should expand vertically as the growth rings will be parallel to the room floor. Cause you are using quaretersawn material you should not have cupping on the drawer sides and even with good dovetails the cupping should be minimal once the drawer in finished.

Did ROb Cosman have the chest of drawers at the Ottawa show for others too try?

Cheers

Dan Clermont

Re: You Forgot one thing...

Todd Hughes

>when asked by VTandy what a piston fitted drawer was Thom D responded they were drawers that pushed "against a cushion of air. You fee a little suction or compression of air when opening or closing"...Rob Crossman said that "pistons needs rings to get this kind of fit"....

So while I realised they were not completely airtight just like I figure you don't really think the drawers are "Pistons" [though that is what they are being called] I did take the drawers to be very closely fitted from the above and other postings so that you can feel the air push in as you close the drawer....close enough to airtight I think, eh?....Todd

Re: Somehow the numbers don't work

Denis Ch�nard, Orl�ans, Ont.

>Dan, the drawer sides will move in both directions. Quartersawn stock is used because it moves proportionally less in that direction (up and down for the drawer sides). But it does move in the other direction too, which here is the tangential plane, where there's more proportional movement (about twice as much for maple). I assumed no cupping in my example to keep things simpler.

Just for the sake of example, assume the drawer side to be 12" high by 12" wide by whatever length the drawer is deep. If quartersawn stock is used then the side would be 1/8" taller and 1/4" wider when MC goes from 6 to 12%.

As for RC's chest, he had one at the show, so I assume it was the one you're talking about. I didn't check for "piston fit", I was too much in awe with all the fine joinery in that chest...

DC

Re: You Forgot one thing...

Stephen in Ottawa

>> Wouldn't it be better to build just a well fitted drawer and not worry about making it an airtight "Piston Fitted drawer?....Todd

Well, at what point does the definition of a "well fitted" drawer end and a so called "piston fitted" drawer begin? I'd consider a "piston fitted" drawer a very well fitted drawer. What I'm trying to say is that the tolerances that are acceptable to one person may not be so to another. Some may consider an absolute precise fit worthwhile, some may not. If one doesn't consider it worth their while, they just won't do it. Nothing wrong with that, but there's nothing wrong with the rest wanting to know how to do it either!

- Stephen

Thanks, Dan

Martin from Granbury

>

Re: Somehow the numbers don't work

Dan Clermont in Burnaby

>HI Denis,

YOur numbers seem high to me on a 12" wide quartersawn drawer (1/4"). Maybe Ottawa has a higher moisture swing compared to Vancouver and our wet winters/dry summers. If I do understand correctly if one builds a 36" wide drawer you need to leave a minimum of 3/4" of space all around even if you use quartersawn maple.

The set of drawers Rob had at the woodshow spent time in England and was damaged in transport by Air Canada. Rob took that set of drawers with him overseas and all over Canada during the woodshow.

I can't look into the numbers you quoted right now as I have a deadline at work but think they are a little high for a drawer with a bottom made out of a floating panel

Cheers

Dan C

Re: Thanks, Dan

Dan Clermont in Burnaby

>No problem Martin. Sounds like you might be entitled to free shipping if you buy a fourth. Don't know if that only applies to Canada or not.

Can't speak for the Peter's video or the M & T video but the Drawer video was great.

Cheers

Dan Clermont

LN describes the dvd as piston fit drawers

Pam Niedermayer - Austin, TX

>

Data, 200years of data on this issue

W. Tindall

>As good fortune would have it I found myself at daughter's in Shelbyville today. She took us to Wakefield-Scearce antique store. I had the opportunity to inspect about 3 doz things with drawers, mostly English, made between late 1700's to mid 1800's and priced $5K to 15K. Essentially all had thin drawer sides of 1/4ed oak and the tiniest dovetails I have ever seen.

Only one piece, a desk, had been fitted with a piston fit drawer, as defined as a 0.00X" tolerance fit. the rest had been fitted 1/64" to 1/16 with 1/32 a median fit(approx). The "piston fit" drawer that still worked did indeed slide sweetly. I can see the attraction to this fit. The rest in this desk had a bind (see below).

As I suspected the potential long term problem with a high tolerance fit is not seasonal expansion; it is stress induced case and drawer deformation. It is a good assumption that these pieces were made well to begin with, ie the case was square. By examination of the drawer reveals I could tell that most drawers and/ or frames had deformed with age, age being 100-200+ years. Also, to my surprise drawer sides had deformed as was the case with the desk. All but one drawer had a bind somewhere in its travel in spite of their 1/4ed sides.

In conclusion, I don't believe that fitting drawers to 0.00X" tolerance will be a good idea over the life of something the size of a chest of drawers or desk because of stress induced case frame and/ or drawer deformation. Except in the cases where there was excessive drawer wear, the large drawers fitted with about 1/32" gap appprox. all worked well in spite of case/drawer deformation.

And, I found a chest with a paneled side that I could peek into. A simple 1/2" high strip nailed to the top of the drawer runner provided lateral support. This not very tall rubbing surface did not induce excessive drawer side wear as I feared it might.

not so over the long haul, see post below

W. Tindall

>

English antiques in US

Ed Mulligan, Cape Cod

>Bill -

Thanks for the post. The truth is somewhere in your numbers. One caveat on your research; in my experience English furniture that is brought to the US has a high moisture content on arrival. As it acclimates to the dryer condition in the US it shrinks and deforms.

If the English drawers show wear marks on their sides, like the photo I posted, it implies they were fitted to close tolerances and any current looseness might be from the change in climate.

Ed

Re: each piece shipped with a well tuned plane

HC Sakman @ Victoria, BC

>That drawer side looks like pine. Is it?

Chico...

drawer side

Ed Mulligan, Cape Cod

>Chico -

The front and bottom appear to be pine; I see latewood 'grain' lines on those parts characteristic of pine. The sides have no visible grain lines and are very soft. I'm guessing basswood but I have very little experience with it or with other softwood candidates. I've talked myself out of tulip poplar as a possibility because the sides are much softer than that species.

Ed

other factors besides mositure change openings

bill tindall

>Using the chest of drawers I have just assembled as an example...48" high chest, 8" drawer with a hypothetical and generous, by piston-fit standards, 0.01" drawer clearance...... The case only needs to rack(twist) 1/16" about its height, that is 0.07 degrees, to close this gap to 0 and the drawer sticks. This is the problem I saw with the antiques I examined. I could see that the drawer sides were no longer orthogonal with the dividers in most cases.

My chest has 12, 6 front and 6 in back, dovetailed dividers. I would judge that well braced. Yet, with but a modest force applied to a corner, this corner moves (chest racks) enough to detect. I presume this to be well over 1/16".

In use, what might cause a case to be deformed this tiny amount and there-by close a tight drawer clearance to zero? Sitting on a floor other than perfectly level would do it over time, expecially a loaded chest of drawers or desk. Moisture changes could, in spite of the arguments of the stability of long grain frame members. Moisture changes are not usually uniform throughout a piece and in any case they will act in complex ways on all the parts of a desk or chest. The rigors of moving the case could deform it. In fact, it is hard to imagine a 48" tall case not deforming more than 1/16" about its height in a 100 years of use. These antiques confirm the hypothesis that over time things the size of chests of drawers and desks will not maintain drawer opening tolerances to 0.01", never mine the piston-fitting tolerances of less than this amount.

I don't know the ideal practical drawer gap for a piece like I am making. I do know that a chest with sticky drawers will likely not be useful (get tossed) while one with less than piston fit will remain acceptable. I suspect the ideal gap is near 1/32 judging from the antiques I examined. A large drawer with an approx. 1/32 gap works fine by my standards. Who would attempt to close it by pushing one corner, loaded? I hope to be able to examine other old pieces to develop a consensus of what has survived the test of time and that will be the standard I build to using the many helpful techniques described in the video.