Strength of Scarf Joints

HomeBuiltAirplanes.com

Help Support HomeBuiltAirplanes.com:

bobm4360

Member
Joined
Mar 13, 2009
Messages
19
Location
Oak Harbor, WA
Very interesting article in Sep/Oct issue of "Wooden Boat" magazine. It covers scarf length, and the method of test. As it's written for boat builders , the materials are oak & epoxy. This is the first material I've seen on actual testing of scarf joints, but there may be some info hidden away in the old NACA reports and Forest Product Laboratory reports.
 

karmarepair

Well-Known Member
Joined
Jan 13, 2011
Messages
77
Location
United States
No data, but AC 43.13 has a lot of "best practices" . 10 to 1 thickness to scarf length minimum, and up to 15 to 1, and they typically call for tapered plywood patches over the ends of the joins. https://www.faa.gov/documentlibrary/media/advisory_circular/ac_43.13-1b_w-chg1.pdf I've cut them by hand (for boats) with a simple jig, but next time I'm building a sled for my bandsaw. Oak is tough to glue, btw.

Didn't find anything from FPL, but they only index about the last 20 years of their work it seems.

But I DID find the WoodenBoat article posted for free https://www.woodenboat.com/strength-scarf-joints . and there is a forum follow-up discussion http://forum.woodenboat.com/showthread.php?256690-The-Strength-of-Scarf-Joints

Short story: don't bother with anything less than 12 to 1, and 15 to 1 more better.
 
Last edited:

mcrae0104

Well-Known Member
HBA Supporter
Log Member
Joined
Oct 27, 2009
Messages
3,219
Location
BJC
Old FPL info from 1920 (?). Not that these joints should be emulated, but it is interesting.

image.jpg
image.jpg
image.jpg
 

blane.c

Well-Known Member
HBA Supporter
Joined
Jun 27, 2015
Messages
3,862
Location
capital district NY
Scarf joints are usually a minimum like for instance 15 to 1? So for example 1/8" plywood to scarf a joint wouldn't it make more sense to go 16 to 1, that is 2" long than some weird fraction out of 15?
 

mullacharjak

Well-Known Member
Joined
Nov 5, 2010
Messages
152
Location
Peshawar Pakistan
I want to scarf the wing spar flanges for a minimax as I cant get good wood in those lengths. I see the grain pattern as depicted in A B and C. I was curious to know which of them is the best on the glued surface. A and B pattern is very difficult to obtain.
C is very easy to get.Would it make any difference with this pattern on the glued surface?
 

Attachments

TFF

Well-Known Member
Joined
Apr 28, 2010
Messages
12,580
Location
Memphis, TN
You want the wood to look somewhat like C from the end and B along the length. If you want to splice a spar, it can’t be a simple one cut scarf. You have to bolster the sides with wood extending past the cut on each side. There is a minimum distance a splice should be away from fittings. AC43-13 has guidelines to splicing spars safely. You essentially have to think like a repair off the bat if you can’t get stock in your length.
 

Aviacs

Well-Known Member
Joined
Oct 21, 2019
Messages
85
The point of the diagram is that you do not want to mix A & B (1/2 on each side) in a wide joint.

A or flat sawn lumber shrinks and expands more across the width and less in thickness.
It will also cup more with changes in RH.
B or Quartersawn lumber shrinks and expands less across the width, and more in thickness
It will also lay closer to flat over a wider range of RH.

True QS will have the grain within about 12 -15 deg of vertical. (actually perpendicular/ "vertical grade" is best) Rift, up to 45deg from vertical will have mixed characteristics, but is better than A. In any even, the configurations should not be mixed.

Mix A & B together, and the joint is fighting itself with every change in relative humidity (RH).
Sometimes multiple events per 24 hr period. Depending on the adhesive & joint prep, delam is a possible eventual prospect. Note that as the material changes in width (with mixed grain assembly), one side will always be trying to be wider than the other. Perhaps more insidiously, the change in thickness of each piece in a scarf joint also causes a change in the relative length of the pieces. There are significant stresses both across and along the axis of the joint.

With "modern" adhesives (Phenol-resorcinal, or very good epoxy) what will actually happen over time is a gradual slight spread or shrinkage of one compared to the other, and micro cracks caused by compression set. This may or may not be actually deleterious to design strength, but it certainly adds "risk" when building.

B is the best configuration for a spar because it is more stable, and the thickness of the wing changes less with variable RH.

What C shows, is that if the grain is diagonal when looking at the end of the boards, they should be oriented so that when the scarf is assembled, the grain in each board is parallel with the other's. Not crossing. In the diagram, the lines are "opposite" when laid open, but will be parallel when the joint is closed. It they are done the other way, then the joint will once again be fighting itself with changes in RH.

smt ,<----scarfs lots of wood for long lengths, especially bending laminae. Never scarfed a wing spar, though.
 
Last edited:

Jerry Lytle

Well-Known Member
Joined
Jan 3, 2014
Messages
358
Location
Newport, Oregon
What about commercial finger joints using with epoxy? I have seen numbers floated around that indicate strengths 85% of unscarfed material using commercial glues.
 

Dana

Super Moderator
Staff member
Joined
Apr 3, 2007
Messages
9,127
Location
CT, USA
What about commercial finger joints using with epoxy? I have seen numbers floated around that indicate strengths 85% of unscarfed material using commercial glues.
They do it that way because it saves material, only uses an inch or so instead of a foot or so depending on the length of the joint. But an amateur wouldn't be able to make such a cut accurately enough without specialized tooling.

It'd be an interesting way to splice the individual plies in a laminated spar, if somebody wanted to go into business making laminated spar stock... might be a business opportunity if the supply of aircraft grade spruce continues to dry up. The engineer in me is thinking of a machine to do it...
 

Aviacs

Well-Known Member
Joined
Oct 21, 2019
Messages
85
What about commercial finger joints using with epoxy? I have seen numbers floated around that indicate strengths 85% of unscarfed material using commercial glues.
I have a couple books on adhesive bonding. The one i reference most for wood is _Adhesive Bonding of Wood_ by M.L. Selbo; 1978.
It is largely a compilation of USFPL experiments and reports with an emphasis on "modern" methods & adhesives to that date.

In the chapter relating to scarf & finger joints there is a mild discrepancy in that scarfs of 1:12 or greater ratio, are described as potentially reaching 90% "strength" of base material. 1:10 is listed as up to 85%. The "potential" aspect is how well the joint is made and glued with the listed figure presumably being for perfect machining, assembly and glue.

In a further diagram of various scarf & FJ assemblies, the pure scarf is listed as up to 85% _tensile_ strength. A note indicates that at that time, up to nearly full strength was possible with clear samples under lab conditions, but that there was no practical means to attain that commercially.

In that same diagram, finger joints vary from 20% to 70% tensile, depending on configuration. Sharp points are better than blunt points but difficult to machine near perfectly.

Another chart shows absolute tensile strength of some 36 different configurations.
Basically, as slope becomes more acute from 1:6, TS mostly improves for the same pitch.
Beyond a slope of 1:12, there is a regression (TS declines) for pitches less than 7/16"
For pitches of 7/16" & 1/2", TS improves up to slope of 1:16.

If the above is confusing without diagrams: Pitch is the spacing of the teeth. slope is the acuteness of each.
The specific glue used for the tests is not specified.

FWIW, type of glue matters & joint prep for full strength is not trivial.

I've read various studies of finger joints over the years and have not retained the details. My formed impression (prejudice :) ) is that they are not reliably close to a true scarf in strength for aircraft elements that have the forces in tension or bending. Might be fine if always compression. A big factor is how difficult it is to machine finger joints perfectly with perfect sharp points, and glue and clamp them all directions as well. For furniture and typical millwork, 50 - 60% or even less is typically fine - the wood is being used primarily for its bulk, more than absolute strength. (Mouldings, a door stile or jamb, etc, etc) In airplanes the idea is to remove as much mass from every component as possible, and stress what's left to within a small safety factor of it's yield.

smt<-----routinely machines blunt/square finger joints in materials that will be formed into flat arches, for machining, but does not count on them for structural purposes :)
 

mullacharjak

Well-Known Member
Joined
Nov 5, 2010
Messages
152
Location
Peshawar Pakistan
I intend to build the scarf with a 1 inch diameter router bit sliding on a slope.
Would the configuration A or B be the better choice for scarf placement.
Also if similiar wood doublers are added as shown in B would there be any need for a scarf ?
Should the doubler be equal or twice in length to the scarf?
Infact such a doubler is already in place in the top flange of a minimax so it might work.
 

Attachments

Dana

Super Moderator
Staff member
Joined
Apr 3, 2007
Messages
9,127
Location
CT, USA
I intend to build the scarf with a 1 inch diameter router bit sliding on a slope.
Would the configuration A or B be the better choice for scarf placement.
Also if similiar wood doublers are added as shown in B would there be any need for a scarf ?
Should the doubler be equal or twice in length to the scarf?
Infact such a doubler is already in place in the top flange of a minimax so it might work.
Doublers on both sides are always used with a scarf joint.

upload_2019-12-8_7-43-9.png
 

blane.c

Well-Known Member
HBA Supporter
Joined
Jun 27, 2015
Messages
3,862
Location
capital district NY
An old article in "Home Shop Machinist" magazine proved that you could get a flatter surface with a smaller dia. end mill. If there is any tilt in your set up from true the error will be magnified by the dia. of the tool used. 1/2" may be perfectly acceptable in wood or if the setup is perfect but it will be flatter with a 3/8" or less if error exist, more passes though.Annotation 2019-12-08 092802.png
 

mullacharjak

Well-Known Member
Joined
Nov 5, 2010
Messages
152
Location
Peshawar Pakistan
BJC
Thanks. A little story.I had applied for permission to construct the minimax to the local CAA.My application was pending since long being pushed from one desk to the other.In the end they asked me for METHOD by which the aircraft was to be constructed. By then my head was hurting.I simply sent a Pdf copy of 43.13.
I knew he would be dead scared of reading it.I was right.
He signed the permission immediately. indeed 43.13 carries a lot of weight.
 
  • Like
Reactions: BJC

BBerson

Well-Known Member
HBA Supporter
Joined
Dec 16, 2007
Messages
12,979
Location
Port Townsend WA
Scarf is normally a repair method.
For home manufacture using short pieces, I would use standard strip laminations. Lighter than multiple scarf scab side plates.
 
2
Top