Tuesday, January 15, 2013

Yeah, But it's a Dry Heat

I am not a scientist. I think that the scientific method is one of the great achievements of mankind, but, I would never claim that I'd worked it thoroughly enough to deserve the designation. But I do want to know more about what I do, so the experiments march on.

Lately I've been thinking about bending. Perhaps it was having to bend some walnut after a recent bad experience that got the juices flowing. I read literature that remarked on the moisture content of steam as it related to wood bending and the concept that steam piped into a box might be lacking moisture. Of course, heat is the primary element in bending wood, but moisture is the conductor. As I understand the concept, the moisture can drop out of suspension in the steam en route to the box and what ends up making it into the box could be heat, but not as wet as it could be.

So to see if my steam needed wetting or if there was any difference in my bends, I attached a reservoir below my steamer that the hose from the steamer runs to so that the heat must pass through water as it enters the box, ensuring that the moisture content was at it's highest.


When I start up the steamer, the steam boils the water in the jar. So far, the results have been positive enough to warrant further testing.


Here are two bends. The same wood, air dried walnut, steamed for the same amount of time, 90 minutes. The piece that bent didn't raise a single fiber even though the fibers ran out the side more dramatically than the failed bend. And yes, the only difference was that one was steamed with the reservoir and one without. Of course, the successful bend was with the reservoir, and as far as free bending walnut, it's the tightest bend that I've done.


I've done other samples and gotten similar results. One factor that interests me is how this effects long steam times. Imagine that the steam isn't adding enough moisture, this may never be an issue with green wood or short steam times, but for working air dried or even kiln dried, this could make a difference.

Like I said, I'm no scientist, and I'd love to here from you about your ideas or experience. I am planning some tests with kiln dried white oak that I'll post soon.


31 comments:

  1. Wow, that is quite the find! Thank you for sharing, Can't wait to hear more about your experiment!! I built my first steam box las fall, but didn't get a chance to try it out, before the cold winter got here!! MAybe some thing to consider, since I work mainly with kiln dry wood...Unfortunately!

    I'll comme back to see what else you find out!

    Can you explain the way you run the pipe and attached the jar?

    Cheers
    David

    ReplyDelete
  2. That is quite interesting. I'd also say thicker wood also needs to be wetter than thinner wood. But overall, that is a slick and simple way to juice up the steam box. Excellent.

    ReplyDelete
  3. Hi Peter,


    It certainly has been my experience that moisture is needed and not only heat. For example when trying to bend a very dry of 1" ash and a 5/8" strip of kiln-dried walnut I found I was successful after soaking them overnight following a first failed attemp without soaking. I also like to soak the inside of my steam box before heating it up to help keep things moist. Lee Valley's technical booklet as more "scientific" indications on the role of moisture to
    plasticize wood. Free on line here : http://www.leevalley.com/en/html/05F1501ie.pdf. I beleive it was prepared by Micheal Fortune who also has video online on this topic.

    Thanks for your excellent blog.

    Manni




    ReplyDelete
  4. The higher the moisture in the wood itself, the better it will bend. Air dried wood is better than kiln dried because the process of kiln drying wood leaves it more brittle, and I have read it also tends to crystalize resins in the wood which hinder moisture penetration. In any case I had 0 success with kiln dried white oak, but air dried red oak and walnut soaked overnight worked like a charm. It is also possible to heat too long and "cook" the wood in the steamer, rendering it brittle. 1" thick equals 1 hour is the rule of thumb in boatbuilding that I have always seen. Insulating the pipe from the steamer and the box is also not a bad idea to help keep the steam steamy.

    ReplyDelete
  5. patelyzo 2085Pete, if you successfully bend kiln dried white oak I'll be very very pleased with you.
    I thought kiln drying permanently set the lignin in the wood making it difficult to plasticise. I've only read this and not actually attempted a bend with kiln dried timber (having plenty of failures to keep me going already).
    You know what we're dealing with down here so if you succeed, the pubs will be filled!

    ReplyDelete
  6. I have had very good success using two wallpaper type steamers (mine are Earlex) and a 7 x 7 x 48 plywood box for the steamer. I make Boggs side chairs and rocking chairs. The rear legs are 1 5/8" thick at the bend. Steam time varies from 2 hours for walnut or cherry, 3 hours for oak, and 3 1/2 for hickory. I always use air dried wood.

    My understanding is that you not only need heat but lots of moisture, which transmits the heat to the wood better than heat alone. I always preheat the water with an electric tea kettle before putting it into the steamers. With two steamers loaded with boiling water the box gets up to temperature (200°) in about 15 minutes, which is when I start counting the steaming time. If I need to refill the steamers, I always use water preheated to boiling so that the temperature (and steam) in the box doesn't go down. This process has worked extremely well, with almost no failures.

    ReplyDelete
  7. Science will save you like a warm blanket!!!

    ReplyDelete
  8. Thanks for all the comments and information. I am doing my best to put my assumptions aside, which is getting easier as some of my results are blowing them out of the water.
    As for steam times, soaking and bending kiln dried, it's open season as far as I'm concerned

    David,
    to run the pipe, I drilled a hole in the top of the steamer and fed the tuber through to the bottom where the jar is. To attach the jar, Tim Manney came up with a great solution which is to use a ball jar lid, without the center cap. He captured the lid by cutting half circles out of two plywood squares and then screwing them to the bottom of the steamer while clamped around the lid.

    ReplyDelete
  9. Well, Pete you know that I bend with kiln dried stuff so I have lots of thoughts on this stuff. I think I sent you a link to blog about bending the 1-1/4" x 3-5/8" blanks in kiln dried walnut for the chair back that I do.

    I think you should post that PDF from the forestry department from the 50's that talks about steaming that I sent you a few months ago. There are a lot of myths out there about steam bending. I would post it but don't think many would find it from my blog nor do I know how to do the file down load links like you have done with your other plans. I know a lot would benefit from it.

    By the way I totally agree with what you are discussing. I recognized early on that if the piece comes out wet on the surface that it would bend but if it came out dry looking that it would fail.

    One of the major things that I think people fail to realize is that if the steam does not have free flow into and then out of the box it will simple not continue to transfer heat and moisture to the parts. The steam will transfer the heat and immediately drop temp then condense the water in the steam box container. Once that occurs you have a high temp, like 211º, but no steam just hot moist air. Since the heat is brought by the steam and not produced in the box then more steam has to continue to flow into the box in order to keep the steam and heat coming in thus there has to be a way for air/steam to flow out. It can't be sealed tight really at all. But when someone makes a steam box the tendency is the seal it up tight and then all they get is water dripping out the end with high temps and then they can only bend green stuff because it is already "wet".

    Took me a while to figure this out myself but having about 12 years of refrigeration mechanics under my belt helped me appreciate heat and moisture transfer issues and how they play a role in steam bending.

    Would like to discuss more. What are you going to call the new invention? The BongBender? :)

    ReplyDelete
  10. Caleb,
    great insight. I've always used a leaky plywood steamer. Thanks for putting some explanation to the benefits of shoddy box building!

    ReplyDelete
  11. Peter
    I find this discussion most interesting as I am about to build a steam box. I went to a seminar a few years back on steam bending and the instructor emphasized moist steam as very important. He accomplished this by building his “box” out of green PVC sewer pipe, not very romantic. ( the green was used as it can withstand the temperature, the white can't) He dammed up the lower 1/3 of the pipe/box by using silicone to seal in some wood blocks, if I remember right. He used this to support the wooden rack or shelf. The steam came into the “box” at the bottom through a manifold of copper pipe, running the length of the “box” (spreading out the steam, no cold spots) that was laying in a bath of water in the lower 1/3 of the pipe. The rack sits above the water. As in yours, the steam had to go through water to get into the atmosphere inside the box-the difference is that the water is in the box (staying warm) and the steam is released over the length of the steam box through the manifold. As I remember, his steam times, once the box was up to temperature, were dramatically shorter than what you are doing. He was quite adamant that if you over steamed the wood you filled the cells with water bursting the cells like balloons rather than moistening the cell walls making them supple and bendable. As one of the other comments in this post mentioned, the pieces came out looking wet and as they were bent you could watch the moisture leave the surface.
    Tom

    ReplyDelete
  12. I have successfully bent white oak, walnut, red oak, and white ash; all kiln dried, all with minimal but some short grain run out. All had mixed results. My learning curve says one needs to steam for 2x the time you would with green wood )or longer). The white oak will probably require a strap too. One wood that catastrophically failed straight across the grain with a loud report was hickory ( much to my surprise and dismay). Please continue the posts on this subject. I love to hear what others experience. Great idea with the reservoir!

    ReplyDelete
  13. I have gone the extra mile in the past by soaking billets in the utility sink in the laundry room for days ( even weeks) with a brick to hold them under the water. My wife tolerates that, but tends to shoot me the stink eye when I put longer pieces the the bathtub. Talk about soap scum!

    ReplyDelete
  14. I steam 1-1/4" x 3-5/8" kiln dried walnut blanks for 40 minutes before bending. I of course have the box up to temp before placing the piece in or starting the clock. I think all that is needed is to thoroughly heat the part all the way through to 212º. Once that is done I am not sure there is any advantage.

    The steaming of parts for too long has only resulted in failures across the grain, for me. I believe it just weakens the fibers and tears instead of stretching. No proof of that last part but just my experience.

    ReplyDelete
  15. That is I meant to add- "any advantage to steaming longer."

    ReplyDelete
  16. Caleb,
    the steam time is definitely a question. I think that the testing grounds is with kiln and air dried stock, I see no reason to steam green wood longer than it takes to get it up to heat. I did have some luck steaming the walnut, with the reservoir, for an hour. It made the bend but had lots of lifted fibers at the surface. When I did the same thing for 90 minutes, not a fiber lifted. Hmmmmmmm...

    ReplyDelete
  17. You know I think my success has a lot to do with the EMC level here in Houston keeping wood stored out doors at a high MC level. I just looked at a chart yesterday from the forestry department that says Houston RH will keep the EMC between 12% -14% through out the year. Which I am sure is probably much lower than your area.

    Of course when you steam it transfers the heat fastest by the water so if the wood has to absorb less then it would theoretically heat through more quickly. I wonder if the longer steam times allow the wood to absorb more water at the surface softening the wood fibers as well as heat the fibers. I know that there is more to it than the moisture serving to heat the wood. It must be moist enough to be flexible.

    By the way going back to the topic of proper flow of steam through the box... I used to have to steam the blank I mentioned before for 90 minutes to get a good bend. After making some modifications to steam flow the time dropped to 40 minutes, for me.

    I have been doing a lot or reading lately about how steam is used in the the kiln drying processes to condition the wood to dry it evenly. This is critical for woods that are "difficult" to dry. Like white oak and beech, among others. It is giving me a lot to consider about how moisture works on woods for controlling its characteristic and behavior.

    ReplyDelete
  18. Great contribution Caleb. I will play with air flow as a factor. The tough part about doing this "scientifically" is that there are so many variables. I am trying to isolate which factors seem to make the most impact on the potential results. Your and everyone elses input is very helpful.

    ReplyDelete
  19. Correction, (again). I mean my wood EMC level "is probably much higher than yours." to begin with.

    ReplyDelete
  20. Yeah, I love this stuff. Wish I could just get a grant from some institution to study this stuff. Fascinating.

    Anyhow, I know you are doing the research to properly explain things clearly in your book so I hope the "isolation" work is fruitful.

    I better get back to work!

    ReplyDelete
  21. Ok, I'm still working on turning legs that I would actually show people, but heat transfer and hydronics is definately an area that lands in my expertise. Steam is just that its a fluid raised above its boiling point where it changes from a liquid to a gas and will move to an area of lower pressure, and will continue to transport heat until its temerature drops below the point at which it condenses turning back into a liquid. As long as your liquid is boiling you are getting all of the potential moisture at the source, (leaving the steamer).

    In my opinion the reason you are having sucess with the resevoir is because it is acting as a remote distribution point (proximity to the wood being steamed) allowing the wood to be penetrated by steam for a longer period of time rather than warm water.

    So, the two factors I would look at changing to have a different out come would be to.
    1) Insulate your box and hose.
    2) Tighten up The air leakage. Obviously you don't want to overly pressurize your box so maybe istall a pressure gauge with a series of holes and plugs so you can keep the box under slight pressure.

    These two factors will certainly allow steam to penetrate your wood much more agressively. I would also think that soaking the KD wood to increase its moisture content would allow for complete penetration of heat from molecule to molecule.

    If you think about this and Pete you being from NYC might find this interesting. A lot of the skyskrapers in NYC including the empire state building do not have on-site heating systems, they are heated by underground (well insulated, and air tight) steam pipes moving from low to high pressure
    I've read so much from you guys, I'm really happy I was able to write knowledgeably about something.

    Andy

    ReplyDelete
  22. One last thought on this from me (for now :) ). It occurred to me that maybe the two factors that are at most to play with regard to time and temp have to do with the MC at which the piece starts at.

    Lets assume that your blank is exactly like mine but yours is 10% EMC and mine is 14% EMC. Lets assume that by the time a successful bend is possible to be made that the blank reaches a 18% MC and 212º. Yours has to absorb twice as much moisture as mine does to reach that MC. That could be the simplest reason as to why you are steaming twice as long. Seems theoretically sound.

    This all assumes that the temp is constant and moisture transfer is consistent. So the factors are a ratio between time, temp, and MC. I think this may be the reason that discussions about time to steam are all over the place because we are not considering the MC to start.

    Of course the very first factor that should be considered in making a successful bend is that it should aim to have zero run out. That gets you off on the right foot for sure.

    ReplyDelete
  23. Caleb,
    Thanks for all the input. I agree that usually I strive for no grain runout, but that piece of walnut in the picture has a bunch of it, go figure. It made me think that I might be pointing the finger at the wrong cause when bends to fail. After all, if they are destined to fail because of some factor, it should be no surprise that they do so at the fiber runout, but that may not be the culprit. Still searching.
    I heard a quote this weekend that I liked "trust the person who is seeking the answer, but beware the one who claims to have found it"

    ReplyDelete
  24. I have seen a lot of steam box setups from a lot of professional chair maker and woodworkers. I have never heard of this. It seems like such a strange idea that steam could hold more water content. Very interesting, thanks for the information.

    David B.

    ReplyDelete
  25. I think you need to add an adjustable pressure relief valve and try different atm's. However, when you do this you will be heating the wood to a higher temp because it is under pressure. Then again, as the fellow pointed out earlier you have to account for hardened resin content in the kd samples of each species...... Which will also draw you into the grain structure, early vs late wood ratio and on and on blah blah blah. My vote is for broad strokes as to how the different woods behave and preempt it with a note on learning by your failures. Wait, didn't you tell me to lighten up and learn by my failures.......

    ReplyDelete
  26. Thanks for posting this subject on your blog. I just got an Earlex steamer in the mail today to replace a crummy kettle I had previously. Tomorrow I will be build a new steam box as well as my older one is falling apart. I've got a few bends to make as soon as the box is ready so I think I will try each bend with a slightly different set up. I will write back with any results I get.

    -Curt

    ReplyDelete
  27. Hi,
    Any one know where I can get plans or dimensions for making a steambox.
    Regards
    Michael.

    ReplyDelete
  28. Paddywack,
    it depends on what you plan to steam. You should build the box about 1" larger than the workpiece. The goal is to keep the space small to maximize the heating,
    good luck
    Pete

    ReplyDelete
  29. Thanks for the tip Peter, I tried out your setup and had better than normal results bending riven yew, sawn yew, sawn elm and riven oak for arm bows and crest rails. All of it dry wood. Not scientific but I noticed a difference.

    Much appreciated,

    Ted

    ReplyDelete
  30. Pete - I know I'm late to this conversation but, there is a bunch of great insight here. I too believe that the moisture content of the wood pre-steaming, is a major factor. The pressure (moving from high to low) is another major factor.

    If the atmospheric pressure in the box is higher than that at the core of the piece to be bent it seems reasonable to think that the steam will more readily work its way through the material. This would then soften the dried and set lignin. On that same line of thought, if the core started at a higher MC it will allow a more "lubricated pathway" as it were, for the heat/moisture/steam to penetrate deeper in a shorter period of time.

    There is a product called Compwood on the market and the way it is made, I believes supports what everyone is saying. The unique thing is the wood is compressed in length prior to being steamed under pressure. This fractures the lignin as I understand it allowing for the steam/heat/moisture to more easily penetrate the wood. The thing is if the Compwood is kept at 20% MC it can be cold bent for up to a year.

    http://www.flutedbeams.com/buycoldbendhardwood.html

    http://www.popularwoodworking.com/techniques/joinery/extremely-bendable-wood

    Based on your experimentation and recommendations of the readers I am about to build a new model steam box that will incorporate a water bath of which the steam will pass through. Also I will attempt to seal it tight enough to raise the pressure (with relief valve for safety). As far as compressing it, I'll leave that to the pros. To think I started out looking for kiln ideas?!?

    As always Pete thanks for your insight and sharing it with we mere mortals.

    Ray

    ReplyDelete
  31. This comment has been removed by the author.

    ReplyDelete