View Full Version : Why Doesn't This Work?

01-04-2016, 08:07 PM
What am I doing wrong? Am I making an error in how am making the measurement, or, as I suspect, is there something fundamentally wrong with using this technique in trying to do this that I am just not seeing?


I used the compass to transfer the bottom plank joggles and angles directly to the floor stock, thinking that it should fit perfectly. The first time I did it, it fit one side but not the other, so I thought perhaps I hadn't been careful enough with holding the compass vertically. So I did it again, taking more care. The second piece also doesn't fit, as you can see. This time it seems that both sides seem to be inboard of where they need to be, and that the bottom has not come down to the keel, which it should have, as the compass was set to that height difference.

I could do this with a piece of scrap and a tick stick or glued on scraps and transfer those to the final stock but I would like to understand why this approach isn't working for me; ham-handed-ness or a a fundamentally wrong approach.

01-04-2016, 08:38 PM
I would do it the hard way, so don't follow me.....

..But I would take a steel ruler and place it on the centre line of your keel, standing up, then measure out to each plank edge from that datum, recording distance away from it, and height above 0. Angles done with a bevel gauge. Recheck it all, transfer to some scrap, check the fit, transfer to floor stock. Once proven and reliable, discard the scrap step.

No doubt someone here will say how it should be done. But I think you're getting compound error in your joggles. Clinker is a pain to run accross like that if it really needs to fit well, if not, just call it extra drainage.

Edit. Please explain exactly how you are using the compass,,,

01-04-2016, 08:59 PM
I'll let someone else explain how to do it the "right" way but I use a hot glue gun and ply wood scrap and it's right every time.

Martin's suggestion of measuring out to the plank laps from a central datum sounds about right.

Clinton B Chase
01-04-2016, 09:14 PM
The trick with the compass is that it is really hard to keep the compass dead square to the piece as you scribe. Any change in angle changes the scribe.

I always teach my students the easy way: plywood scrap ripped to about 2" and length to suit and hot glue to tack it in place. Scribe the individual ply pieces if you need to. Glue the ply piece together with gussets and a cross spall if needed. Then list it out...sometimes a tap with a hammer snaps the hot glue welds. Remeber with this technique...one side of the ply pattern represents the shape you want...the other side does not.

01-04-2016, 09:27 PM
It is not exactly clear what you are hoping to have in the finished result. Does the floor timber sit on top of the keel leaving a vertical gap on either side? If that's the case, I'd do it the way lupussonic described above. Use the keel top as the datum. There will be a point where a line from the top of the keel intersects the first plank. Mark that point on each side of the timber. Then simply measure up to the first "joggle" and transfer that measurement to the timber (it will be a semicircular mark from the first intersection point. Use the bevel gauge to get the angle and that will tell you where the first "joggle goes. After that it is really just a matter of taking distance and angle measurements from that point out on each side.

Since you have not described how you are doing this, it is pretty hard to say what you might be doing wrong, but from the look of things it may be that you measured the face of each plank, and then transferred that laterally to the timber to locate the next "joggle". It has to be transferred on an angle, otherwise it will put the joggle too far out (which is what it looks like happened). Think of this like dead reckoning navigation. You start at the point where a horizontal line that touches the keel top intersects the plank. And then from that point you dead reckon your way to each of the subsequent "waypoints"...From each point where the direction of the cut line changes, there is both a distance to the next point where it changes, and an angle, so you would scribe a circle (or a portion of a circle) with the correct distance, and then get the angle between the current cut line and the next cut line with a bevel gauge. Where that angled line intersects the circle is the location of the next point.

Like this:


Phil Y
01-05-2016, 06:01 AM
What's the problem? I love thickened epoxy;)

wizbang 13
01-05-2016, 06:21 AM
Yes , Too much brainwork.
24 grit game ova.

01-05-2016, 06:22 AM
Never thought of hot glueing strips together, wow you just saved me 1000 hours of my life Peter. Thanks!

01-05-2016, 07:18 AM
"SCRAPS" of plywood???? There are no "scraps" of any kind of wood in my shop! Seriously, for templates, I use corrugated cardboard.

Ian McColgin
01-05-2016, 07:30 AM
I think the illustration at #5 replicates the OP error.

Make a spiling batten straight at the top and bottom that lays on the keel, arced on each side reasonably close to the planking. Make it wide enough that you can strike the same size arc FROM each data point. The data points should be the outside corner and inside corner of each step. Best to use two different colours, say red for the inside corners and blue for the outside. You'll want a stick as thin as your spiling batten with a 90 degree V in one end to place in each lap, giving something to position the fixed end of you compass against but it will still take a steady hand. Make and label the arcs though hopefully they are all far enough apart that with the two colours you'll not get confused.

I'd transfer to a template to be really sure. Put the spiling batten on the template material and strike two intersecting arcs (same radius as the originals!) from each arc on the spiling batten. There are your data points ready to connect the dots.

Whether you hot glue up a template as in #3 or are making a one-piece template as I suggest, cheapest material is to go to the dump and strip off some door skin from tossed away doors. Basically free bit under 1/16" single ply stuff of good wood.


01-05-2016, 07:37 AM
1/4" luan door skins at Home Despot. $11 a 4x8 sheet.

Ian McColgin
01-05-2016, 07:41 AM
Another good source if you don't have a decent town dump.

01-05-2016, 08:11 AM
You are all wrong, it is the pencil. Only yellow works well in a compass.

Steve McMahon
01-05-2016, 08:42 AM
For any complex shape like that I like to use a large plywood board held in place with a drop of glue gun glue, 1/4" plywood or hardboard, laid on the keel and touching a point on each side. Then make a joggle stick from a scrap of 1/4" thick x 1" wide material, about a foot long, square on one end, a sharp slash cut on the other end. Hold the joggle stick point at each of your marks and trace the square end of the stick onto the plywood, noting which face of the joggle stick is up (mark your joggle stick on both sides).

Once you have all the points clamp your marking board onto another sheet of template material and reverse the process to put the points onto the template. Cut out the full size template and double check fits.

There is a better explanation with a fancier joggle stick than I've used here:

If you google "joggle stick" you will find lots of good stuff.

David Satter
01-05-2016, 08:49 AM
Iv'e been using chipboard , non corrugated cardboard. For pattern making. the stuff is wonderful , comes different thicknesses, the one I use is about thick as a dime. It's stiff , easy to cut. Use that with a can of spray glue, or glue gun. makes great accurate patterns. I use it for knees, stems, transoms. I get 2'x3' sheets. The big sheets are nice because you can transfer the curve of a stem onto one big piece. Dave

Measures Once Cuts Twice
01-05-2016, 09:02 AM
unsure which comment to leave so you get the list:

1). what's not working? that's well within my tolerances....

2). i always leave it a bit big and tall and sneak up on the correct fit (if a grinder="sneaking") then when it fits the bottom mark a line on the 'tall' top and power plane to it (remembering that it will all then be hidden under the floor boards...)
-and i see i could save lots of time with the mock up, glue stick, approach!

3). i love these type of question threads; i is always learning stuff here :-) Thanks!

01-05-2016, 09:51 AM
Iv'e been using chipboard , non corrugated cardboard. For pattern making. the stuff is wonderful , comes different thicknesses, the one I use is about thick as a dime. It's stiff , easy to cut. Use that with a can of spray glue, or glue gun. makes great accurate patterns. I use it for knees, stems, transoms. I get 2'x3' sheets. The big sheets are nice because you can transfer the curve of a stem onto one big piece. Dave

Yep, from an amateur's viewpoint, make a pattern, takes care of the old measure twice, cut once theory.

01-05-2016, 09:58 AM
I'll let someone else explain how to do it the "right" way but I use a hot glue gun and ply wood scrap and it's right every time.

Martin's suggestion of measuring out to the plank laps from a central datum sounds about right.

I use plywood scraps too, but I hold them together with little spring clamps, then remove the assembly from the boat and staple it. I'm sure hot glue would be better.

David Satter
01-05-2016, 10:25 AM
If anyone does this for a living ( I do, a modest one, if your in it for the money...) There's always that time aspect. How much time can I spend on this? It's always a mind game, you want it to come out perfect but you have to stay in budget. I hate making that call saying it's going to cost a little more. If the boat is for your own personal use, you have to say lets get it in the water at some point. Ok my breaks over I have to get back to work.

01-05-2016, 11:13 AM
^ Thats what i do. patterns ensure you dont waste time OR material. I would not go back to a compass.

Gib Etheridge
01-05-2016, 11:51 AM
I would do it just as you've already done Alex, but with 1/4" pine sawn and thickness planed from 2x4 scrap. The first try would never be to the tolerances I prefer, but once it's that close it's easy to put it back in place and scribe it again with a much smaller setting on the compass. I'd probably go at it a third time also.

I used to restore colonial homes where nothing was straight or square or plumb and much was hand hewn and this method always worked the fastest for me.

I would keep the oak a bit too wide (high) and once it fit well to the planking use a short level to mark the top to the right height.

01-05-2016, 12:49 PM
Alex, there is nothing fundamentally wrong with your technique. In an ideal world, using your pencil compass will give you an accurate scribed line to cut to. BUT… as Jim Ledger points out, human error will always get in the way. Not keeping the compass perfectly vertical, not having a super sharp pencil point, letting the stock slip just a little, etc. will all introduce errors that will add up.

By the way, more error can be introduced depending on how one allows for the actual thickness of the stock having to conform to the curvature of the hull. The profile should ideally be scribed/transferred to both sides of the stock to be fit. Then those lines are connected to allow for hull curvature. That means twice as much measuring and marking. And as a consequence, there is twice as much potential error introduced. By this time, the boat builder is tempted to go have a stiff drink and to contemplate the qualities of thickened epoxy.


01-05-2016, 01:16 PM
Thanks all for taking the time to reply. It's great to see the various perspectives and experience.

I might just go back and try re-scribing with the compass set closer and cutting again and see where that gets me. It will only take 10 minutes and the stock is already a write-off.

As for various template methods, tabs or joggle sticks. I know they will work. That's what I did with my bulkheads and they fit pretty closely, first time. Somehow I was hoping to "save" time with the compass. What's the old saying - never enough time to do it right, but always enough time to fix it?

I have lots of doorskin bits left over from the templating I did for my planks, so no excuses really. I am also well aware of the error introduced by stock thickness and hull curvature. That's why I started with this floor in the middle of the boat where there is almost no curvature - the supposedly easy one.

01-05-2016, 01:26 PM
Yeadon's pic is worth a thousand words.

01-05-2016, 01:50 PM
Compass scribing works like this: if you hold the compass perfectly vertical through the scribe and cut to the lines perfectly then, from the stock position when scribed, the piece will fall vertically the span of the compass and fit perfectly. Held consistently at another angle the piece will fit as though moved in place at the angle. This of course ignores bevels, compass angle irregularities, etc. None of us is perfect. We have all probably imagined that tracing with compass square to the stock will work and not disappoint but it is not so.

Gib Etheridge
01-05-2016, 02:16 PM
Using thin stock for the template has the advantage of allowing you to scribe just the "big" face of the piece. Once that fits and it's transferred onto the thicker stock and cut out all you need to do is plane and rasp/chisel the bevels. You can do those by eye if you just take off a little at a time. In my previous post I recommended 1/4" pine, and that will work just fine midships where there is little bevel, but as you get towards the ends where the bevels are greater you will do better to use 1/8".

I don't like using doorskins. They're too weak and brittle and splintery, and they bend too easily.

Ian McColgin
01-05-2016, 02:20 PM
I disagree with Thad a bit. If you don't hold the compass fairly normal to the circle, you could be pushing the point off center and make an inaccurate arc, buy unless you've squeezed the compass legs, you've not changed the distance between the points. That's why when you use a compass to measure distances on a chart, it measures the same whether you lay the compass at a low angle of hold it normal to the chart. But Thad's observation that holding the compass at an angle may induce error is correct.

01-05-2016, 03:58 PM
NO. Holding the compass at the same angle to vertical is correct whatever the angle, the angle being the line between the points not the angle of the body of the compass.

Jay Greer
01-05-2016, 04:29 PM
Spiling the jogs pattern for floors or frames in a clinker built hull is not all that hard once you get a clear picture of it in your mind. Making up a pattern by using a batten sprung to the high faying points of the planks will allow you to bring the pattern into contact with the planks prior to notching the pattern. The point of intersection to the pattern, where the planks touch, should be first marked using a small tri square to the inside of plank lap. Then the thickness of the inner lap at the step is marked off on the pattern. Projecting the lines from each plank lap is then a simple matter. Note that the pattern should be marked from the side that is closest to the midships line in order to allow for the bevel of the floor. If this is a small boat, you can often eliminate the frames or just bend them in to be in contact with the inner high spots of the jogs. Such as can be seen here.
https://im1.shutterfly.com/media/47a6df24b3127cceeb5f78ae7d6000000030O00QYsmrNy5bsQ e3nwg/cC/f%3D0/ls%3D00107990352120160105212505137.JPG/ps%3D50/r%3D0/rx%3D720/ry%3D480/

John Meachen
01-05-2016, 04:34 PM
Thad is absolutely correct.What you absolutely must not do is rotate the compasses,you are translating the shape followed by the point to the area beneath the pencil point and to allow any kind of rotation displaces the pencil line.To illustrate-consider trying to fir a short length of wood inside a piece of fairly large pipe if you kept the legs perpendicular to the surface you will describe an arc smaller than the pipe radius and if you then cut the wood,it will rock when dropped into place.If you maintain the same orientation of the legs,the wood could be cut to a very good fit.

Ian McColgin
01-05-2016, 05:08 PM
I must not be understanding what Thad and others mean. I just took a compass and struck a circle with the compass (centerline up between the legs) normal to the paper. Then I tilted the compass back and forth striking arcs and a whole circle. All I did was make the pencil mark of the original circle darker. The legs remain the same distance apart and thus the radius of the circle is the same.

A different problem comes up drilling a hawse hole in the bow planking of a boat. There the circle of the pipe is piercing the plane of the planking at an angle. If you use a compass to scribe a circle on the outside or inside of the planking, you'll get a marking suitable for a round hole normal to the planking at that point. Since that's almost never the shape you want, you might drill a fine pilot hole angled as you want it, then make a sort of compass of a rod that fits the pilot hole and a bar out holding the scribe. You can then mark the hole's shape pushing the rod in or pulling it up to keep the scribe in contact with the plank. Perhaps something like this phenomenon is what others have in mind.

John Meachen
01-05-2016, 06:06 PM
If my explanation fails to make things clear,Chappelle describes the technique in "Boatbuilding" and there is an illustration which accompanies it.

Gib Etheridge
01-05-2016, 06:08 PM
I think you were agreeing with each other. You both agreed that the line described by the 2 points must be vertical, and you, Ian, pointed out that it didn't matter whether the compass was leaned to the left or the right, so long as the lead at the end of the one leg was directly above the point at the end of the other leg. Si?

I think what you're describing for the hawse hole is best accomplished by cutting out a round or an elliptical hole then using a rasp to bevel and round the edges.

01-05-2016, 06:19 PM
I believe that there are a million ways to skin a cat. To do it the way you want should work fine also, and should give you the right piece the first time. Go to your local LOG HOME builder, He can tell you where to buy the correct tool. Google 'log scribe' and amongst the miriad of scribes, you'll spot it. I've used it to scribe logs, to duplicate the shape of the log underneath to the perpendicular (or any stationary angle) log on top, both sides, both ends, roll the log over, cut it, roll it back and drop it into place. The tool looks just like what you have there but it has bubble levels on it. I know nothing about boat building tho.

Is that the answer?

01-05-2016, 07:24 PM
Don't discard the piece, epoxy on a top piece.
Or it might work further forward.
Always start with the widest one for this reason.
You are on the right track using a scribe vertically.
You need to learn this, looks like you have a good number to do.
Make your stock wider by a 1/2" or so to allow for fitting and cut the top to size when the bottom fits.
It will get easier as you get the "knack".
You can do three floors in the time it would take to make one template shown in #20! :)
My two centavos, anyway.

01-06-2016, 07:33 PM
So, I had another go at it, with the same original piece, with the compass set to the remaining difference, as suggested by Gib Etheridge, and lo and behold, it worked, or, close enough anyway:


I was also able to salvage the first piece for another location - it'll just need a little more goodge to make the fit.

I haven't decided whether I'll use the same 2-step technique for the remaining floors. I could use this technique on pattern stock and make any mistakes on that before transferring to the good stuff. It is quick, I'll say that for it.

Gib Etheridge
01-06-2016, 08:24 PM
That's a lot closer, for sure. The third try is what zeros it right in, that and the bevels, but it's hard to be really accurate when the surface you're scribing to mandates that you need to bevel the piece. That's why
I suggest that you try the pattern one time as well, it works better because it's so thin that the not yet shaped bevel doesn't hold it up away from the hull. I'm certain that once you've tried it you'll find it gets the best results in the shortest amount of time. I don't see why it would be any different for you compared to what works for me.