Replies

1. 
   bob_barnett | Mar 16, 2009 03:00pm | #1

   Thanks for the post; that is very interesting.  This is the first time I have ever seen any information on spring back.  The equations are much simpler than I would have expected to see.  I was a little surprised that they did not account for differences in material properties of the different species of wood.  Each of the equations will give a slightly different answer.  What is your experience with the accuracy of the equations?  From my testing of wood and comparing it to deflection equations, the results are not as uniform as the books would lead you to believe. 

    

   Bob

1. 
   SgianDubh | Mar 16, 2009 03:34pm | #2

   Bob, the formulae give, in my experience, only partially useful results. This is why I mentioned in the first paragraph that I often resort to intuitive but informed guesswork.

   Did you notice that none of the formulae make an allowance for laminate thickness? Yet I know that an 18 mm thick bend made of 3 X 6 mm thick laminates generally springs back more than an 18 mm thick bend made of 6 X 3 mm thick laminates. Slainte. richardjonesfurniture.com

1. 
   bob_barnett | Mar 16, 2009 07:07pm | #3

   <!----><!---->I<!----> will have to admit when <!---->I<!----> am not playing in my shop or not sailing (what <!---->I<!----> am currently doing until June) <!---->I<!----> am a structural engineer.  <!---->I<!----> hate to admit being an engineer on the list because people think you are some sort of weird know it all.  The things <!---->I<!----> noticed about the equations were the lack of the (E) Modulus of Elasticity and (<!---->I<!---->) Moment of <!---->I<!---->nertia.  Modulus of Elasticity is a measure of how stiff the wood is.  Moment of <!---->I<!---->nertia is normally a function of the depth cubed.  So, if you double the thickness of the material it should be eight times as hard to bend and <!---->I<!----> would think more spring back.  <!---->I<!----> have also dealt with equations enough to know that some things just cancel out and others are close enough to a constant you just plug a value in for that variable.  Even with all that, <!---->I<!----> agree with you that the thickness should be in there somewhere. 

   The reason this subject interest me is <!---->I<!----> enjoy building wooden boats.  Finding a tree with the perfect bend is not as easy as cutting up strips of wood and epoxying them together.  One of the issues <!---->I<!----> always have to deal with is spring back.  Since <!---->I<!----> don’t build the same boat twice <!---->I<!----> never know how much to compensate for on spring back.  <!---->I<!----> will try your equations next time, add a little judgment, and see how well <!---->I<!----> do.  Thanks again for sharing your information in spring back. 

   <!----><!----> <!---->

   Bob

    

   PS  I love the work y'all do at Richard Jones Funiture

1. 
   RickL | Mar 16, 2009 08:23pm | #4

   "Springback can be predicted with the formula y = x/n2. The ratio of springback to the original deflection depends only on the number of laminations. The ratio does not depend on the properties or thickness of the wood or the geometry of the curved form. Thus, for two plies, the springback is one quarter of the initial deflection, or one ninth for three plies and one sixteenth for four plies. "

   I don't recal seeing his formula 1 but Formula 2 I first saw in FWW over 20 years ago. It's a ballpark average so there is no elasticity or inertia information to go with it. It does seem to work on the occcasions we've used it in our shop and using a stiffer glue gives less springback. In fact thinner plies and a stiff glue gives almost none or negligible springback

1. 
   WillGeorge | Mar 17, 2009 04:14am | #12

   Maybe I should keep quiet as I never want to start a 'fight' with anybody. However, when I read anything on the technical side, my little brain fills with all sorts of questions. That said I will continue.

   In fact thinner plies and a stiff glue gives almost none or negligible springback...

   I guess it depends on how you define 'springback'

   On occasion I will make a Archery Bow. I am NOT a Bower. I usually miss the whole target in front of and hit the one to my right.

   I only make them because it is fun to see if they sort of work. To my standards, not to a true Bower...

   All the Bows I have ever made were laminated from various woods. Usually Hickory for the center and Ash as the face and back. Sometimes some fancy wood for the face and back. Fancy woods 'for looks only', not function. I have no idea what I am doing except trying some woodworking I find fun to do...

   When I make a Bow I think of Springback in two ways. As in the way the bow 'stays in shape' after it is removed from the gluing form under some normal weather extremes. AND I WANT ALOT of Springback when I release the Arrow!

   I have use many types of glue and different thicknesses of different types of wood.

   I have never seen any difference with the type of glue used except for a Bow that has to 'see' extreme moisture (out in the rain all day) and back into it's normal home.

   I guess all that I am saying is the glue and size of the laminations never seemed to matter once the bow/wood was 'tillered'. But then again it is hard to 'tiller' a hunk of furniture and have it look uniform.

   Sorry, I had to...

2. 
   1woodmaster | Mar 16, 2009 09:29pm | #5

   Hi Bob,

   I also am an engineer.  When I ask a fellow woodworker about springback, I get some pretty vague answers.  This sounds like a great article for fine woodworking.

   There are so many variables though.  It would take a great deal of emperical data to determine the target results.  Types of glues, material species, variations within species like type of cut (plain, 1/4sawn, etc), moisture content, kiln dried or air dried, number of laminations, moment of inertias, variations in radii. The discussion would have to be limited to rectangular cross sections, non-tapered and non-compound bends. 

   I monitor the Knots forum for items of interest that may make good articles and I think this is a good one.  There must be some info out there on this subject.  I would hate to go through all of the work if someone has already done it. 

   I just googled the subject and found a few formulas but nothing really helpful.  I would think that with a thoughtful plan of what info one is looking for and a roadmap of how to get it, this subject would be doable.  I plan to write "Ask The Experts" to see if FW has extensive info on this topic.  If not, and I suspect not, I will develop an experiment and data aquisition plan.  Once that is complete, I would like to send it to you for review to see if there may be something that you can see that I overlooked.  The experiment may take a year or more in my spare time, but I'm patient and time flies when you're having fun.

   Scott

1. 
   bob_barnett | Mar 16, 2009 09:50pm | #6

   Scott

   <!----><!----> <!---->

   <!----><!---->I<!----> would be more than happy to review your material.  <!---->I<!----> was going to look into the subject more myself when <!---->I<!----> got back to the office.  <!---->I<!----> have been trying to think how this issue could be modeled in an analysis program and <!---->I<!----> don’t have a clue.  This has to be a horizontal shear issue along with slippage or creep of the joint.  We are living on a boat right now and it will be sometime this summer before <!---->I<!----> am back in the office around all my books.  There maybe something on this subject in the “Wooden Boat” magazines archives or publications. 

   <!----> <!---->

   Bob

2. 
   SgianDubh | Mar 16, 2009 10:01pm | #7

   Interestingly, everyone seems to be assuming that springback is bound to happen. It doesn't. I have seen what I call "spring-in", ie, the curve gets tighter after it is released from the form. It is rare, but I've seen it, and not just with laminated bends; I've also seen it in solid single pieces of steam bent wood.

   This is why I take all these formulae with a good dose of salt. I just wonder how scientific one can get when you are dealing with wood-- the stuff is so darned unpredictable.

   I put that short wee article up on my website simply because I thought people might like to know about them. I can't recall where I got the formulae from now. I think I've had one in my possession for nearly thirty years. If I knew where they came from I could acknowledge the source.  Slainte.richardjonesfurniture.com

3. 
   1woodmaster | Mar 17, 2009 03:17am | #10

   <!---->You are so lucky to be living on a boat.  In what area do you live? I lived in SanFrancisco for 3 years and was considering the boat idea then.  Really cool!

    

   I tried to go into ask the experts in FW and could not find a way to ask the experts.  Have you tried it?

    

   I've read a few entries after yours.  It seems that people (not just in these entries but in general) feel that wood is unpredictable and that accurate predictions of behavior are not possible.  It's probably my engineering  mentality, but I don't agree with this.  You just need to identify the variables and make them part of the equations.  For instance: There is a big difference between steam dried wood and air dried wood.  Because of what the drying process does to the cell structure of the wood, steam drying yeilds a more dimensionally stable but weaker product.  These variables just need to be treated as two different materials.  I really believe that this process of developing a formula for spring back is simple but very labor and time intensive.

    

   :You're probably reeling in your dinner and sporting a glass of wine on the boat every day.  I also read occassionally read the wooden boat magazine.  It's fun to dream.

    

   Scott<!---->

4. 
   bob_barnett | Mar 17, 2009 05:23am | #14

   We are just on the boat for a few months.  We are just taking some time to roam around south <!----><!---->Florida<!----> before heading back to our homeport in <!----><!---->Pensacola<!----><!---->.  Now if you want to daydream how about a boat big enough to have a table saw on board. 

   I tend to agree that wood is very variable.  I purchased a couple of 2x12 a few years ago to load my 4 wheeler in the back of the truck.  One of the boards would deflect a couple of inches more than the other one as I drove up the boards.  They both had the same dimensions and were both labeled as #2 but there was a remarkable difference in the deflections.  There are just many things we can’t take into effect when we try to predict behavior of members.  Steel probably has the most predictable behavior and wood the least predictable behavior.  Every piece of wood is unique which is what I love about working with it.  Wood from each tree will behave differently, they will perform similar with similar loads but there will always be differences.  I have been able to test a lot of beams and reviewed a lot of post tensioning and cambering data over my career.  If you get deflection test within 15% to 20% you are doing pretty good.  Some of the codes recognize this in their allowable tolerances.  The allowable tolerance in cambering a steel beam is 3/8” plus 1/8” for every 10 feet.  So if you have a 30 ft long beam your acceptable camber is between 0” and 1 ½”.  With that much acceptable error, I guess it is good we have large safety factors. 

   <!----><!----> <!---->

   <!---->bob<!---->

5. 
   SteveSchoene | Mar 17, 2009 01:20pm | #15

   The Forest Product Laboratory is working on ways of apply probability distributions and statistical analysis to the properties of wood.  Oh my. 

6. 
   1woodmaster | Mar 18, 2009 02:03am | #16

   Hey Steve,

   How do you know this?  Do you have a contact at forest products?  It would be nice to know what data they have obtained or to know when they might have some results on wood properties.

   Scott

7. 
   SteveSchoene | Mar 18, 2009 05:22am | #18

   If you google for Forest Products Laboratory, and then go to their web pages, there are links to their links of publications, and also links to current research areas.  You can see what they are working on by looking at either set of lists.  They have a very active research program, and much of the work is published on line, though some is seriously technical. 

8. 
   1woodmaster | Mar 18, 2009 05:23am | #19

   Steve,

   Thank you very much! This is a great find for me.  I really appreciate it.

   Scott

9. 
   1woodmaster | Mar 18, 2009 02:18am | #17

   Hi Bob,

   Are you retired? I was wondering how it is that you can travel and be away for 3 months?

   My email address is [email protected]. If you would like to chat about common interests it would be fun to hear from you.  I also have Skype set up with a webcam.  If you have Skype you can reach me through a Skype search on my email address. 

   As long as I'm on the subject, Skype is a very nice free program that I use in conjunction with a webcam to do visual conference calls with friends, relatives and fellow woodworkers around the world.  It runs through the internet and is all free.  My cousin in Norway has a wireless webcam.  Last week when we were talking, he walked around the house showing me how nice the dining table looked for the dinner party they were having and the new fish stew recipie cooking in the kitchen. The only cost is the initial cost of my webcam which was about $35.  If you havn't tried it, its a great way to stay in touch with friends and meet new people all around the world.

   Scott

10. 
    KeithNewton | Mar 17, 2009 05:01am | #13

    Well Bob, As you consider all of these variables, just keep in mind that each of those variable properties are working against the next ply and the next one, and etc. The tension of one is compressing the next, and basically it just boils down to more plys being better than fewer. I have never had any spring-in, as Richard mentions when using laminated bends. I normally use epoxy, thickened with colloidal silica. However I have had some steam bent parts spring in to a shorter R. On those occasions, I attributed it to starting to get grain wrinkling on the inside, due to pushing near the limits for a thickness. Since the wood would then be shrinking due to the moisture in the steam being lost, the deviation of the grain could explain why there would be a shortening on the inside of the bend. Besides, any shrinkage across the grain, with no change in the longitudinal axis should cause a shortening radius due to the new memory sets while the wood is wet. Yet the cross grain contraction occurs after as the wood dries. When the cells dry, the length of the inside arc and outside arc have not changed in length, yet they have changed in radius, and grown closer. If their ends stay even, which it does. This across grain contraction should cause a turning inward.

3. 
   WillGeorge | Mar 17, 2009 03:40am | #11

   I<!----> hate to admit being an engineer on the list because people think you are some sort of weird know it all.

   I have worked with many different types of Engineers  Sorry, no structural Engineers. Mostly Mechanical, Electrical, Chemical. The only weird one had a PH.D. something in Physics. Optical, Lasers and Sound. He was a 'true' nut and a very close friend.

   I found almost all to be very down to Earth types UNTIL you I said something that needed correcting or you disagreed with them.. :>)

   At one time my direct Boss was a VP of the company I worked for and she had a PH.D in Mechanical Engineering. (Not sure what her degree was but I know she had all kinds of diplomas and awards all over her office wall. She was someting to work for.I would hand her my reports on 'things' she asked for and I would get them back 'graded' like the Teachers I had in School... All kinds of red ink and references to books or white papers I should read... She was funny! I never knew if she was serious or just pokin' fun at me. She sure knew her 'stuff' though. I guess I did OK anyway because I worked for her about 8 years and always got good 'reviews'.

   I always found that 'most folks', educated or not, that knew what they were doing never tried to impress you. OK, maybe a bit to show off their skills (I see nothing wrong with that) Although, you may have to tell them "I have to go home tonight" when they talked on some subject....

    

    

2. 
   dustymctevish | Mar 16, 2009 10:15pm | #8

   Do you take into consideration the type glues used in the laminations or do you believe it negligible?  I’ve been leery about some glues tendencies to creep under stress.

    

   DustyMc

    

1. 
   SgianDubh | Mar 16, 2009 11:14pm | #9

   I simply avoid creepy glue, ie, PVA types. I'd say 95% of any laminate work I've ever undertaken was with urea formaldehyde glue which dries rigid, has gap filling ability and is waterproof. It's got all the necessary characteristics for good laminate work.

   I've done a little laminate work with polyurethane glue which is okay so long as you can get adequate pressure because the expanding foam can separate laminations. I have used hide glue, but in this case I've only ever used liquid hide glue that I can recall. I might have used hot hide glue for the job back in the 1970s, but if I did it was probably a small and fairly straightforward job. I've never used epoxy resin for the job that I can remember-- I might have done, but if I have I've forgotten it. Slainte.richardjonesfurniture.com