I have a project on my hands where I have to create a smooth curve. I
plan to sandwich a number of MDF pieces together to achieve depth.
However, for structure, I also need to sandwich a number of 2x4's
(Pine) in there (facing edges will alternate between pine and MDF).
Imperfections and joints will then be bondo'd to achieve a
consistently smooth surface and later painted. I'm hoping to reach
the end result of a smooth contemporary piece that looks like it came
out of a mold.

Is this plan bound for cracks in bondo/paint as humidity changes and
the piece is moved around? Would I be much better off with hardwood?
I'm aware that wood breathes, but I have no clue how much. I plan to
use wood screws and "General Project" glue (brown caulk-like
container) for assembly.

Thanks for any and all advice!

On May 8, 11:22 am, fourrings wrote:
I have a project on my hands where I have to create a smooth curve. I
plan to sandwich a number of MDF pieces together to achieve depth.
However, for structure, I also need to sandwich a number of 2x4's
(Pine) in there (facing edges will alternate between pine and MDF).
Imperfections and joints will then be bondo'd to achieve a
consistently smooth surface and later painted. I'm hoping to reach
the end result of a smooth contemporary piece that looks like it came
out of a mold.

Is this plan bound for cracks in bondo/paint as humidity changes and
the piece is moved around? Would I be much better off with hardwood?
I'm aware that wood breathes, but I have no clue how much. I plan to
use wood screws and "General Project" glue (brown caulk-like
container) for assembly.

Thanks for any and all advice!

Have you considered bending wood? It is a product made for cabinet
makers and is in stock or can be ordered from plywood dealers. Cabinet
shops use it to build curved casework such as reception desks. I've
never done it, but I presume you can laminate it with P-Lam to achieve
the look you want. I think the bondo idea is asking for trouble.
Joe G

more information please.

how much strength do you need? MDF can be pretty strong, in some
applications.

what are you making?

is your curve more than 8' long? if so, the face ply can be one piece-
no need for filler beyond fastener holes.

Ah, somebody brave enough to attack the problem at the root I'd
greatly appreciate some feedback on the design. The curve is def more
than 8' long. The table needs to be 40" tall with a 50" span - so
figure 130" total (depth of 25")

Imagine this glass table ( http://www.umodern.com/fimages/970.jpg )
but painted black and consistent thickness of ~6". The top will need
to be semi-hollow as I will be sinking audio equipment into it
(~150lbs worth). Because the table will be so top-heavy and possibly
prone to tipping, I'm reconsidering option 1 in favor of option 2:

1. Make box frame out of 2x4 for legs and top. Then use metal
brackets to attach top to legs. Finally, cover frame in 1/2 or 3/4"
MDF.

2. Top frame remains a box from 2x4's. But legs are made out of many
sandwiched MDF pieces already with the curve. For structure, I will
need to embed a couple 2x4's in the sandwich and use same metal
brackets to attach legs to top.

fourrings wrote:

Thanks for any and all advice!


How about bending plywood?

You can get it with many different facings, and it's pre-kerfed to
facilitate bends. Check with your local hardwood dealer, as it's
usually a special order item. Plywood wholesalers usually stock it, so
the special order won't take long to get.

You MIGHT be able to order it from a home center, as they often deal
with the local plywood wholesalers, but you'd need to know exactly what
you want, as the home center won't.

fourrings wrote:


Imagine this glass table ( http://www.umodern.com/fimages/970.jpg )
but painted black and consistent thickness of ~6". The top will need
to be semi-hollow as I will be sinking audio equipment into it
(~150lbs worth).

Google "torsion box" and make the whole thing from plywood. Standard
3/4" stuff for the frame, and bending plywood for sheeting. The shape
can be constructed of one-piece plywood girders attached to
crossmembers, wich form the tosion box. The whole shebang then gets
sheeted with bending plywood, chreating a strong, relatively light
structure. Think "flying model airplane" or "plywood boat".

Forget the 2x4's and MDF. On another note, 2x4's are not properly
dried for furniture, so you'd have to store them an awfully long time
before you use them.

On May 8, 1:23 pm, fourrings wrote:
Ah, somebody brave enough to attack the problem at the root I'd
greatly appreciate some feedback on the design. The curve is def more
than 8' long. The table needs to be 40" tall with a 50" span - so
figure 130" total (depth of 25")

Imagine this glass table (http://www.umodern.com/fimages/970.jpg)
but painted black and consistent thickness of ~6". The top will need
to be semi-hollow as I will be sinking audio equipment into it
(~150lbs worth). Because the table will be so top-heavy and possibly
prone to tipping, I'm reconsidering option 1 in favor of option 2:

1. Make box frame out of 2x4 for legs and top. Then use metal
brackets to attach top to legs. Finally, cover frame in 1/2 or 3/4"
MDF.

2. Top frame remains a box from 2x4's. But legs are made out of many
sandwiched MDF pieces already with the curve. For structure, I will
need to embed a couple 2x4's in the sandwich and use same metal
brackets to attach legs to top.

Many carcases similar to your description are made by constructing a
framework of 3/4" plywood. The plywood pieces are cut to the cross-
sectional shape you want and then layed out in the desired pattern.
Its something like the wings of a balsa wood airplane model. Then the
bending wood is applied, giving you the smooth curve. This becomes the
substrate to lay up the P-Lam. This method is very strong and should
hold your equipment
Joe G

Google "torsion box" and make the whole thing from plywood. Standard
3/4" stuff for the frame, and bending plywood for sheeting. The shape
can be constructed of one-piece plywood girders attached to
crossmembers, wich form the tosion box. The whole shebang then gets
sheeted with bending plywood, chreating a strong, relatively light
structure. Think "flying model airplane" or "plywood boat".

Forget the 2x4's and MDF. On another note, 2x4's are not properly
dried for furniture, so you'd have to store them an awfully long time
before you use them.

Bending plywood looks interesting and a possibility. Though, I was
hoping to accomplish my goal with commonly available material. Plus,
isn't the surface fairly rough? I would need something for the frame
though as the top span will deflect too much under weight. Is
hardwood a sufficient alternative to Pine in this case?

fourrings wrote:


Bending plywood looks interesting and a possibility. Though, I was
hoping to accomplish my goal with commonly available material. Plus,
isn't the surface fairly rough?

Not if you buy it with a smooth surface. You can even get a smooth MDF
or laminate surface, if that's what you want.


I would need something for the frame
though as the top span will deflect too much under weight.

That couldn't be more wrong. Too bad you didn't actually read up on
torsion boxes. 8^(

Do you realize that 747's and major bridges are built with torsion boxes?

Do you realize that 747's and major bridges are built with torsion boxes?

Irrelevant, but long ago I actually designed and built a bridge based
on the torsion design. This was done out of balsa using crazy glue.
At the competition, not a single stick fractured but the bridge
twisted and snapped at the glue joints :-/

I improperly pictured the proposed solution when I replied. Now that
I better understand the cross-sectional design, I think it's an
excellent idea. So essentially, I take two sheets of (any particular
wood type?) 3/4" plywood, cut out U shapes from each sheet to form the
front and the rear cross section. Then lay bending plywood on top and
bottom? What kind of load will this design be able to support with
only a 6"-wide cross section? Will the corner in the bend be strong
enough and not tear apart without reinforcement?

What are my options for joining the cross-sections together? I will
need to center three 50lbs (each) components inside the table top.
Each will be about L15"xW15" and need to be placed/removed from top of
table. This means that whatever flooring I use for the table top
span, will need to be sturdy.

fourrings wrote:
So essentially, I take two sheets of (any particular
wood type?) 3/4" plywood, cut out U shapes from each sheet to form the
front and the rear cross section. Then lay bending plywood on top and
bottom? What kind of load will this design be able to support with
only a 6"-wide cross section? Will the corner in the bend be strong
enough and not tear apart without reinforcement?

Use more than two girders. Think of your recessed gear on top, as
you'll need to work around it.


What are my options for joining the cross-sections together?

How are rails held together in any torsion box? Notched crossmembers of
the same material.

I will
need to center three 50lbs (each) components inside the table top.
Each will be about L15"xW15" and need to be placed/removed from top of
table. This means that whatever flooring I use for the table top
span, will need to be sturdy.

3/4" cabinet grade plywood (the kind with lots of plies) is very strong.
You could sheet the center section with that, and bending plywood on
the curve, with internal stringers to bridge the seams.

Got it. I'm thinking two cross-sections, three inches or so apart, on
either edge of the table (four total. maybe one in the center as
well? - though it will be interrupted shortly after the curve).
Reinforce (glue and screws ok?) with cross-members as seen in this
image - http://www.raygirling.com/images/closeup.jpg. Am I close?

Can I use screws as well as nails to secure the cross-member to cross-
sections? And where do I notch the other side of the cross-section?
Exactly across or shifted a bit? Please excuse my lack of
terminology, I'm learning as I go. So in the above image, if I have
another horizontal cross-section to join, would I shift the notch or
not?

THEN can I use Bondo??? Or will P-Lam take it's place.

You might try looking on diynet.com, for a "Freeform Furniture" episode
where AMY (host) makes some curved nesting tables using bending plywood.
You may be able to get some ideas about your tables construction. If
not...oh well, Amy's kinda cute.

here is a link to the site...
http://www.diynetwork.com/diy/shows_...09,00.htmlgood luck.Mike"fourrings" wrote in ooglegroups.com... Got it. I'm thinking two cross-sections, three inches or so apart, on either edge of the table (four total. maybe one in the center as well? - though it will be interrupted shortly after the curve). Reinforce (glue and screws ok?) with cross-members as seen in this image - http://www.raygirling.com/images/closeup.jpg. Am I close? Can I use screws as well as nails to secure the cross-member to cross- sections? And where do I notch the other side of the cross-section? Exactly across or shifted a bit? Please excuse my lack of terminology, I'm learning as I go. So in the above image, if I have another horizontal cross-section to join, would I shift the notch or not? THEN can I use Bondo??? Or will P-Lam take it's place.

Sorry about the link...try this

http://www.diynetwork.com/diy/shows_..._40709,00.html


Mike

"MikeG" wrote

You might try looking on diynet.com, for a "Freeform Furniture" episode
where AMY (host) makes some curved nesting tables using bending plywood.
You may be able to get some ideas about your tables construction. If
not...oh well, Amy's kinda cute.


Amy is cute. But she makes atrocious furniture.

If they gave out awards for ugly furniture, she would be a constant winner.

transfer the weight of the components to the top skin. think of a tray
that the electronics sit in, bonded to the underside of the top skin.
you can bond it to the bottom skin too, if you want, but the important
loading is to the top.

the inturrupted rib isn't helping much. a lighter rib right on either
side of the component tray will probably be better.

screws into the edge of MDF are pretty much useless.

don't notch the long ribs (the ones that turn the corner and bear the
weight of the top) very much. don't make them from MDF either. the
transverse ribs can float between the long ones. it will be convenient
for assembly to fasten them into a frame before applying the skin, but
not necessary for the strength of the finished article. try to either
limit yourself to fasteners that don't displace a lot of wood (like
finish nails) or have correct pilot holes for the screws you use.

shifting the notches (staggering the transverse ribs) is fine, and
will help avoid weak spots lining up with each other. better is to not
notch.

this assembly is similar to a hollow core door. in a door, the
function of the ribs is taken up by some lightweight, more-or-less
continuous material like foam or cardboard. you need to bear weight
against the panel, so you will need a thicker top skin and some ribs
running the long way. other than that, though, it can be a lot like a
hollow core door. the cross ribs can be replaced with shaped blocks of
foam, if you want. the bottom skin, the harder one to assemble, can be
thinner that the top. if you can, the cool way to make the ribs is to
either steam bend or laminate hardwood.

Thanks to all for great advice! I've shifted my plan and will be
using the suggested torsion box design with 3/4 ply. I just need to
iron out a few more kinks before I start.

Bridgerfafc -- I've been convinced that I won't be using MDF for this
project. Unfortunately, I won't be able to transfer the weight to the
top skin since the components don't have an edge that they can rest
on. What I was thinking though is that instead of doing the entire
bottom-center skin out of 3/4 ply, I could use the thin bendable ply
for the bottom skin and then use separate 3/4 ply sections under each
component for support. Each section would be the width/height of the
individual component and would be screwed into the edge through the
long ribs and the cross-members. Though I'm curious about the
strength of this method. Now that I re-read your suggestion for the
third time, is this exactly what you're suggesting?

So don't bother with the center long rib? This means that I will have
the ribs spaced as such: 0"--5"--20"--25". Sufficient span for
bendable plywood?

If I don't have to, I'd rather not notch the long ribs. Especially
since this will weaken them to some extent. So, standard butt-joint
with glue and wood screws (with pilot hole) it is.

Another puzzle that I'm trying to figure out is this: I want to round-
off the corner edges for a sleek look. However, how do I do this if
the top/bottom skins will be screwed onto the edge of the long ribs?
The only solution I can come up with is to double up on outside ribs
(doesn't have to be a continuous piece since it's not for support -
per say) and use one rib for attaching the skin, and the other for the
rounded edge. Any better suggestions?

I won't be able to transfer the weight to the
top skin since the components don't have an edge that they can rest
on. What I was thinking though is that instead of doing the entire
bottom-center skin out of 3/4 ply, I could use the thin bendable ply
for the bottom skin and then use separate 3/4 ply sections under each
component for support. Each section would be the width/height of the
individual component and would be screwed into the edge through the
long ribs and the cross-members. Though I'm curious about the
strength of this method. Now that I re-read your suggestion for the
third time, is this exactly what you're suggesting?

it's close, and it will work, if you make suru the 3/4" bottom panels
are well attached 4 sides to a rib and that that rib is well bonded to
the top.



So don't bother with the center long rib? This means that I will have
the ribs spaced as such: 0"--5"--20"--25". Sufficient span for
bendable plywood?

the center long rib will still help stiffen the curve from top to leg,
so it's not a total waste. make sure it terminates at the top into a
cross rib, and that that connection and the ones between that cross
rib and the long ribs IT terminates into are strong joints.


If I don't have to, I'd rather not notch the long ribs. Especially
since this will weaken them to some extent. So, standard butt-joint
with glue and wood screws (with pilot hole) it is.

Another puzzle that I'm trying to figure out is this: I want to round-
off the corner edges for a sleek look. However, how do I do this if
the top/bottom skins will be screwed onto the edge of the long ribs?
The only solution I can come up with is to double up on outside ribs
(doesn't have to be a continuous piece since it's not for support -
per say) and use one rib for attaching the skin, and the other for the
rounded edge. Any better suggestions?

sounds right.
build the frame, apply the skins a bit oversize, trim, apply the outer
rib a bit oversize, trim and round over.

wrote:


it's close, and it will work, if you make suru the 3/4" bottom panels
are well attached 4 sides to a rib and that that rib is well bonded to
the top.


As Bridger points out, the entire skin is structural. If properly glued
to the sringers, it'll be incredibly strong.

Good quality plywood stringers will be far stronger around the corners
than hardwood. Plywood has many plies with alternating grain direction
to carry the load. A hardwood stringer would have some directions that
aren't as strong as the others. The stability of the plywood will also
help the joints and bonds last.

it's close, and it will work, if you make suru the 3/4" bottom panels
are well attached 4 sides to a rib and that that rib is well bonded to
the top.

So screws driven between the plies will not compromise plywood's
strength to any significant extent and will not separate the plies
from the weight, right?

Good quality plywood stringers will be far stronger around the corners
than hardwood. Plywood has many plies with alternating grain direction
to carry the load. A hardwood stringer would have some directions that
aren't as strong as the others. The stability of the plywood will also
help the joints and bonds last.

I'm very glad I asked this group before I started the project! The
torsion box design you guys suggested is MUCH better than the amateur
one that was brewing in my head.

Is Bondo still my friend to fill in fastener holes, joints, and other
voids before painting?

fourrings wrote:

Is Bondo still my friend to fill in fastener holes, joints, and other
voids before painting?


It's good stuff, but avoid using it where there might be flex.

If you work carefully, you can probably get away with very little filler.

It's good stuff, but avoid using it where there might be flex.
If you work carefully, you can probably get away with very little filler.

I'd imagine there would be some flex where the skin meets the ribs.
Also, since I won't be able to get a long-enough skin, I'll have to
stitch it half way. This would also have to be filled. If not bondo,
what should I use?

fourrings wrote:
It's good stuff, but avoid using it where there might be flex.
If you work carefully, you can probably get away with very little filler.

I'd imagine there would be some flex where the skin meets the ribs.
Also, since I won't be able to get a long-enough skin, I'll have to
stitch it half way. This would also have to be filled. If not bondo,
what should I use?


You'd probably be OK.

I was thinking more along the length of the stringers.

If you properly glue the skin to the structure, you'll have an amazingly
strong and rigid structure. There shouldn't be movement at the splices,
because you will have placed crossmembers under them that pick up both
sheets. Each area within the table becomes a box girder.

Did you ever think of wrapping the whole thing in plastic laminate? You
could use one sheet on the top, running from the bottom of one leg, over
the top, to the bottom of the other, and again on the bottom side.

One more thought...

As a design element, you could leave the top and bottom skins slightly
wide, creating two interesting, parallel lines, when the table is viewed
from either side.

As a design element, you could leave the top and bottom skins slightly
wide, creating two interesting, parallel lines, when the table is viewed
from either side.

That's definitely an idea. Though I will have a control plate flush-
mounted in one rib, which will make access less convenient under an
over-sized skin.

As for laminate, I figured that's one more specialty product that will
require prior experience to apply properly. I figured if I mess up on
painting, I can always sand down the troubled spots. This will also
allow me to have a seamless surface that I'm looking for.

fourrings wrote:

That's definitely an idea. Though I will have a control plate flush-
mounted in one rib, which will make access less convenient under an
over-sized skin.

Don't forget that holes in the rib will compromise strength. A
"doubler" rib forming another box behind the controls will help.

Don't forget that holes in the rib will compromise strength. A
"doubler" rib forming another box behind the controls will help.

There will be five 1/4" holes (and 1" apart) in a horizontal
arrangement. You think that will be a problem?

I just came back from HD. Found three possibilities for 3/4"
plywood: Sandeply, Oak, and Birch. Sandeply is $32 per sheet and
others are $40. Which would be ideal for the ribs and cross-sections?

fourrings wrote:
Don't forget that holes in the rib will compromise strength. A
"doubler" rib forming another box behind the controls will help.

There will be five 1/4" holes (and 1" apart) in a horizontal
arrangement. You think that will be a problem?

Probably not. Round holes help the stress travel right by. I was
thinking of a square box placed into the rib. In a load carrying part,
round holes are better than square holes with round corners, which are
better than square holes.


I just came back from HD. Found three possibilities for 3/4"
plywood: Sandeply, Oak, and Birch. Sandeply is $32 per sheet and
others are $40. Which would be ideal for the ribs and cross-sections?

The choice with the highest number of plies and fewest voids. Ideally,
you'd get cabinet grade Baltic or Finnish birch plywood. There are
versions of this with "paint grade" face veneers that aren't that
expensive, but you may need to order it at the Pro Desk.

More plies equals a stiffer rib and more edge wood for the skin to glue
up with.

Probably not. Round holes help the stress travel right by. I was
thinking of a square box placed into the rib. In a load carrying part,
round holes are better than square holes with round corners, which are
better than square holes.

It's not exactly a box, just a metal plate that will be flush-fit into
the "fascia' rib. But only the five 1/4" holes will ever penetrate
the load-carrying rib. Good to know about the effects of various hole
types though.

The choice with the highest number of plies and fewest voids. Ideally,
you'd get cabinet grade Baltic or Finnish birch plywood. There are
versions of this with "paint grade" face veneers that aren't that
expensive, but you may need to order it at the Pro Desk.

More plies equals a stiffer rib and more edge wood for the skin to glue up with.

Sandeply has seven plies, but the surface is not MDF-smooth
unfortunately. I figure that I could fill any voids with Bondo and
work the surface with an OS to get it smooth? Provided sandeply is
strong enough for this application. If I can get away with $32 ply
instead of a $50+ special one without any side-effects, that'd be
great. I'll have to call some local plywood distributors to see what
I can find though.

(At the rate I mention Bondo, I should be doing auto body work... I
just keep reading all the praises of its use on wood. I also expect
to fix a lot of mistakes in this project).

fourrings wrote:
Provided sandeply is
strong enough for this application. If I can get away with $32 ply
instead of a $50+ special one without any side-effects, that'd be
great. I'll have to call some local plywood distributors to see what
I can find though.

I don't have any experience with Sandeply.

Be aware that most home centers can get cabinet grade plywood and some
of it defaults to 5' x 5' sheets. 5x10's are often available, if you
need continuous parts longer than 5'.

B A R R Y wrote in news:VgJ0i.3161$RX.727
@newssvr11.news.prodigy.net:

fourrings wrote:
Provided sandeply is
strong enough for this application. If I can get away with $32 ply
instead of a $50+ special one without any side-effects, that'd be
great. I'll have to call some local plywood distributors to see what
I can find though.

I don't have any experience with Sandeply.

Be aware that most home centers can get cabinet grade plywood and some
of it defaults to 5' x 5' sheets. 5x10's are often available, if you
need continuous parts longer than 5'.


Stepping in here, from the other side of the country...

I've used Sandeply, sourced from my local, pro plywood supplier. It's not
what you think of when you say baltic birch, or wasn't in my case. I made
bookcases out of it for the sister-in-law school teacher, and painted it
with latex semi-gloss. It worked, but just barely.

Were I to build something that takes the level of thought that fourrings is
putting into this job, I'd go with Finn birch or Appleply from one of the
pro suppliers. Spend the extra $25. Stronger. Cleaner. More stable.

YMMV.

Patriarch

Good quality plywood stringers will be far stronger around the corners
than hardwood. Plywood has many plies with alternating grain direction
to carry the load. A hardwood stringer would have some directions that
aren't as strong as the others. The stability of the plywood will also
help the joints and bonds last.


unless the hardwood is laminated or steam bent to the curve. then it
would likely be much stronger than a sawn plywood rib.

So screws driven between the plies will not compromise plywood's
strength to any significant extent and will not separate the plies
from the weight, right?

you'll have to test that hypothesis on YOUR plywood with YOUR screws
and YOUR skins. test pretty agressively, too- bang it around some to
make sure.

me, I'd probably start with 1/4" crown x 1-1/2" staples, partly
because it's what I have and partly because the staples displace so
little wood.

As for laminate, I figured that's one more specialty product that will
require prior experience to apply properly.

yes, but it's not too bad. there are a few kind of important things
to know, but it ain't rocket science. also, because of it's density it
makes an especially suitable surface for your application. it will add
stiffness and strength to your torsion box.




I figured if I mess up on
painting, I can always sand down the troubled spots. This will also
allow me to have a seamless surface that I'm looking for.



Laminate is available in some pretty long sizes, and it presents a
professional looking finished surface from the start. getting a really
true looking surface is a lot about doing the prep work, which is:
bondo/sand/test/bondo/sand/etc....

wrote:

unless the hardwood is laminated or steam bent to the curve. then it
would likely be much stronger than a sawn plywood rib.

Agreed!

I have a stack of ash that will become a toboggan this year to enter
this event:

http://www.camdensnowbowl.com/tobogganNat.cfm

Thanks for the heads-up, Patriarch. I'll skip on Sandeply. I've
called a few suppliers yesterday and gathered that imported Birch and
Appleply is hard to come by. And the default 5x5' will not work for
me. So if I don't find anything, I might have to settle for HD Birch.

Thanks again for everyone's help! I now have all the info I need to
start.