Glen-l plans misaligned (Monte Carlo)?

[bikemaniac]

I have done the following so far: I obtained the Monte Carlo runabout blueprints from Glen-L and got them scanned. Then I imported the images to Autocad and finally I drew the lines on top. Somehow I noticed that some lines were not completely perpendicular so I went for the original blueprints with a stainless steel ruler to check for alignment using trigonometry. Unfortunately my guesses were right. Here is one example:

I have one long centerline and perpendicular to that centerline, I have a "set up level reference line". Those two have to be 90,00 degrees. They are not! They are 89,76 degrees (measured in Autocad) and by checking the blueprint I can see that the lines are not perpendicular (pencil, calculator, ruler, pythagoras). Actually the offset is measured to 4 milimeters. Thats a lot to my opinion!!!

Well, what should I do? Should I use the centerline or the "set up level reference line" as the correct line for the frame? Can anybody please help me out?

Regards, Lucas Jensen

[cs]

If you are trying to scale off of a drawing you can end up scaleing in mistakes. Same thing with scanning them in. Paper drawings can tend to "move" and this could be what is causing the error.

I would make this line perpindicular when I layed it out.

Chad

[mmd]

Pretty good example of why lofting is important, I'd say. Lay out your base, reference, and station lines square & parallel on the lofting floor, draw the boat's lines from the table of offsets, and everything else should fall into place.

[bikemaniac]

@ cs & mmd:

I'm not doing any scaling: Everything is perfectly 1:1. This has been checked and is 100% accurate. The paper has not moved during scanning - what I'm saying is, the error I see in Autocad is found in the real world blueprint too!!!. The scan and Autocad is not my problem - my problem is the original blueprint.

As of lofting: I'm sorry but I don't know what you mean about lofting. All the lines are there 1:1 - lofting is not necessary - Glen-L has done all the hard work for me :-). I would happily lay out my base and reference - the problem is just that those two lines are NOT 100% perpendicular.

Thanks in advance, Lucas

[Uncle Duke]

Chad is perfectly correct - paper "moves". In the cartographic/mapping world (where I spent many years) this is a known and serious issue, and special ('expensive') papers are used where needed to reduce hygroscopic expansion. I don't imagine that blueprint paper is any more stable than map paper, so scanning and expecting perfection is probably an exercise in frustration.

The "Bible" of cartography ("Map Projections - A Working Manual", USGS #1395) states the problem like this:
"... a more serious problems results from the fact that maps are commonly plotted and printed on paper, which is dimensionally unstable. Typical map paper can expand over 1 percent with a 60 percent increase in atmospheric humidity, and the expansion coefficient varies considerably in different directions on the same sheet".

edited to add: folds in the paper can also create dimensional errors...

[cs]

I'm not an expert in lofting, but lofting involves taking the x,y,z cordinate and laying it out on the shop floor. You can than fair by eye to remove any hard spots.

Anytime lines are put to paper there is room for error, whether you draw it in by hand or you send it from your AutoCad software to the plotter. Paper moves with temp, and line width can play into it also.

I would imagine that the line was meant to be perp., but somewhere in transfering the lines from either a program or a hand drawn drawing that the line got off a little.

I'm not familar with the plans you have or how they are laid out. I would attempt to make the correction and make it look right.

Chad

[RonW]

Bikermaniac- Since I have built from glen-l plans and am familar with their system, what you have is not a set of plans, but full size patterns and instructions. The patterns are made to be transferred by means of pins, or the tracing paper that glen-l sells, they are not and will not copy properly with out the paper movement as stated above.
I doubt if the lofting tables are even included with, shall we say patterns, instead of plans.

Glen-l designed the patterns so you have everything you need to build one boat, not make copies and build a dozen boats.
Pretty slick huh, hey this is the real world and if you want to build more then one boat, then order more then one set of patterns, or make a set of plywood templates.

[Thorne]

Isn't this why some designers provide patterns on mylar or other non-paper material? This issue is a common one from what I hear.

[mmd]

OK, so you know for sure that you don't have to loft, the patterns are full size (must be a big piece of paper), the paper hasn't stretched (unlikely), and the draftsman made an error. This is what you do:

1.) Call Glen-L and inform them of their error - they will surely want to know this and make corrections to their master drawing.
2.) If there is a baseline, determine if it is normal to the centreline.
3.) If the answer to (2) is "yes", use this as your datum line for set-up.
4.) If the answer to (2) is "no", draw a new "level reference line" that is truly normal to the centreline and originating from the point that the old reference line crosses the centreline.
5.) To check that the new "level reference line" is accurate for the building mold setup, measure the distances from the sheer point on each mold line to centreline parallel to your new reference line. On a flat clean surface (such as a piece of plywood) draw a baseline and perpendicular to this, at 4" intervals, draw a line the length of each measured mold half-breadth, in sequence from transom to stem. Take a batten and spring it from tip to tip of each line and draw a line from first to last line. If your new reference line is accurate, this line will be a fair smooth line that just touches the tip of each half-breadth line; if the curved line doesn't touch the half-breadth lines or intersects one or two a few millimetres below the tip, you still have a problem.
6.) If the process described in (5) yields an unfair line, repeat the process using measurements taken parallel to the supposedly crooked reference line (it could be that the centreline on the pattern is inaccurate and the Glen-L draftsman drew the reference line accurately but did not correct the centreline... unlikely, but possibe).
7.) If (5) and (6) do not indicate a fair layout of the molds, contact Glen-L and ask for their advice.

Do not use AutoCAD for the above layout exercise because 1.) you want a real-world batten to spring the line and, 2.) AutoCAD does not necessarily create fair curves with the polyline utility. Which brings me to my final question: If Glen-L has provided you with full-size patterns to use to construct your building molds, why did you scan it and run the risk of introducing errors due to scanner head scaling and paper distortion due to tension created by scanner feed rollers?

Finally, even though you do not need it for this project, it would aid your understanding of how to build boats if you learn a bit about lofting. There are several books available for minimal cost, one (a good one) sold by our host. Such information is certainly worth the investment.

[JimD]

I purchased plans for three glen-l boats and got around to building two of them. The boat I did not build was Amigo. It came with both full size patterns on paper and a table of offsets. The instructions said to stick to one or the other but don't mix and match. I found the patterns for the two boats I did build to be plenty accurate enough to trust.

[dmede]

@ cs & mmd:

All the lines are there 1:1 - lofting is not necessary - Glen-L has done all the hard work for me :-)


Lofting can most certainly be necessary, even with full size plans. Unless you got your plans on mylar or something similar there is a great chance that the paper has "moved" as another poster suggested. Meaning it has streached or deformed in some way. I had the same problem with a fullsize bulkhead drawing on a recent canoe build. The problem in that case was that the drawing was simply a scaled up print of the original scaled down plans, which might look correct at 1:12 but not 1:1.

If you have offsets for the plans you should be using them instead and the drawings as a reference only.

dave

[George Roberts]

For all practical purposes 89.76 degrees is 90 degrees.

So many ways to solve problems.

Make the vertical centerline the true centerline. Make perpendiculars through the zero crossings.

4mm = .18". Hard to build that close.

[JimD]

Persuant to my previous post...

I was in Alberta at the time. Continuously dry climate almost year round, very little cause for the paper to move.

[Tom Robb]

If you have patterns, use them to build the boat.
If you want to make it as dead accurate as possible and they sent the table of offsets, loft it and build from the lofting.
It's only a boat, not the replacement mirror for the Hubble telescope.
Trying to reverse engineer, or whatever it was that you were trying to do with Autocad, is not the way to go. Copying and scanning only introduces more errors.
If the advice you get is not the advice you want, then why ask?

[Bruce Hooke]

George raises what I think might be a good point. It is not clear what distance the 4mm error occurs over, but a 4mm error in many boatbuilding situations is within normal tolerances. It is easy in AutoCAD to get too wrapped up in precision. As the saying goes, machinest work to the nearest 0.001, cabinetmakers to the nearest 32nd and boatbuilders to the nearest boat (or something like that -- actually in cabinetmaking I usually find it is more a matter of working to the nearest scribe mark).

It might help if we knew why you were trying to go through all these hoops to get the drawings into AutoCAD. If you are trying to get set up to use some sort of automated system to mill out the parts then I can see why you want great precision, but the lesson here may be that full size paper templates are not really the best starting point for such an approach.

[jzeigler]

My two cents. It may be the process of "scanning" itself by the Glen-L folks. I've built many half models and before scanners were available, I relied on the local Photo Repro shop to do the scans. They were in fact, photos enlarged maybe a couple three times. If they thought they could get away with it, they only took one shot and enlarged it. The distortion to the edges of the photo from the camera lens created a "drift" somewhat like what one sees in a gnomonic projection of a chart. I built the half model of Australia 2 to imperfect plans before I realized the distortion (too much sheer, keel wings at wrong angle)
So lofting will find and eliminate any problems.

[Ron Carter]

Two posts, no experience, asks for advise and then argues with the advise given. Ya just can't make some people happy. Either build it to the patterns or loft it. As said above it isn't a precision instrument, it's a hole in the water!

[cs]

I don't think he is really arguing the point as much as defending what he thinks is right.

Chad

[Steve Miller]

I built a Glen-L Utility skiff last year. I used a combination of the patterns and the dimensions on the drawings to build the boat. I was able to work to a tolerance of 1/8th inch and was pretty darn proud of that as a fairly novice builder (3 boats now). The patterns are indeed full size but given as 1/2 of each frame and transom. The stem pattern was full size.

I am not sure how you could easily use the pattern sheet to layout individual frame pieces without cutting the patterns out but that won't work since Glen-L nests and overlaps the patterns on the sheet to get it all to fit. So I transfered the patterns to a master template piece of plywood and used that along with the plans to layout and assemble frames. The bottom of the Utility is an arc so rather than use the half patterns I drew the arc on my pattern board full size which was a good warm up for the two deck beams which were not given on the patterns but need to be cut to fit the hull.

Glen-L clearly marks the base reference line on the plans and each frame pattern so setting up the form and mounting the frames, stem and transom was pretty straightforward.

I could have built the boat from the plans alone since they gave all the critical dimensions of the frames, bottom and deck arc, transom rake and so on. No need to loft at least on the Utility and no table of offsets was included anyway.

[bikemaniac]

I think the confusion started me not knowing the difference between a pattern and a plan. Forgive me, I'm from Denmark and English is not my mother language/tongue. So what I have drafted so far in Autocad are the PATTERNS which are 1:1 FULL SCALE. Sorry for the linguistic mistake.

The main reason why I'm drawing the patterns in Autocad is for backup: If the originals get damaged, wet, burnt or whatever I will always be able to plot a new one!

I got no figures from Glen-L to loft the boat.

To all who don't like Autocad and scanning: I have already drafted one pattern in Autocad. It goes as follows: The pattern is scanned and I obtain a huge .tif-file. This tif-file is imported into Autocad and is set as a background. Now I just start tracing directly on top of the background drawing making everything vector graphics. I can assure you that no mistake is done here. I have printed one pattern and I put it on top of the original blueprint pattern while holding it against a large window (using tape). I can tell you the Autocad drawing was dead on!! Everything matched the orinal pattern 100% accurately.


Let me try to explain the error more precisely:

Imagine the following:
1. Draw a line exactly 1000 mm long (this will be your centerline).
2. At the very end of this line draw another line 90,00 degrees (also 1000 mm long) with respect to the first line (this will be your set up level line).
3. At the intersection between the to lines at the very end of the two lines, draw another line 89,76 degrees (also 1000 mm long) with respect to the first line (=centerline).
4. There will now be a gap between the last two lines. At the far end (=away from the intersection of all 3 lines) the distance between the ends of the two lines will be exactly 4,18 mm (calculated by means of tangens/trigonometry).

What I just explained above is exactly what I'm seeing on my pattern - and exactly that worries me.

@George Roberts:
You are saying that I should use the centreline as the primary line and align the set-up line at 90,00 degrees with respect to the former. Does the rest of the group agree on that statement?

Happy building, Lucas

[htom]

I think what you may have missed is that in making the-shipped-to-you blueprint from the orginal drawing there can be stretching of the paper (or film, if it was drawn on that), and that that error has propogated to your print. Probably errors, in fact, slightly different on each sheet. This error strikes me as huge but I'm used to thinking in terms of masks for making integrated circuits.

[Steve Miller]

Maybe I am not getting this but the centerline of the boat is a fore and aft line that should intersect the middle of the transom, frames and stem. The set up level reference line is marked on the plans and pattern for each frame and the transom and would intersect the centerline at 90 degrees. Actually the two lines MUST intersect at 90 degrees. The set up level reference line on my Glen-L is also the top of the building form. This line gives you a reference for the vertical mounting of each piece on the form. As an example the photo below shows the centerline (the chaulk line and the vertical pencil line on the frame). The set up level reference line is also faintly visible on the frame. For this specfic frame the set up level reference line is 2" below the top of the bottom frame piece so the frame was mounted lower than the base line on the form thus you can see that the top of the 2x4's meet the pencil line. If your patterns do not have the centerline and Set Up Level Ref Line meeting at 90 degrees then just correct that when you build the frames. It is VERY likely that they do not as explained by all the posts above on the problems with paper patterns.

http://www.hevanet.com/kg7pv/Utility/image15.jpg

[George Roberts]

bikemaniac ---

Yes. That is how I would do it.

[bikemaniac]

@Steve Miller.

Now I'm confused. I took your photo and added some lines to show:

http://photos.yahoo.com/faksgnisner/

The fat lines show what Glen-L mean by centerline/setup line.

Lucas

[RonW]

There is nothing to be confused about.
Just set things up at 90degrees, and quit worrying about a 1/4 of one degree of error on paper.
You are worrying too much about nothing.....

[Steve Miller]

The set up level reference line in your photo shown in yellow is correct but your yellow line creates an optical illusion since it overlays the photo not the actual form. It may clear it up for you if you shorten the yellow line to just overlay the pencil line on the frame itself and not the form pieces. The set up level reference line on the frame in the photo is lined up with the top of the horizonal 2 x 4's that run fore and aft NOT the vertical 2 x 4's that are the form legs as shown by your yellow line overlay. Otherwise your photo is correct. Note that the set up level reference line may be in a different place on each frame depending on the shape of the hull. As RonW said - just build them at 90 degrees.

But if you are still confused then try this: (you will have to do this for each frame and the transom to build the boat anyway)

1- draw two lines at 90 degrees on a piece of plywood or paper making a big +. Lay one of your frame half patterns on it on the left side of the intersection.
2 - align the paper 1/2 pattern so that the set up level reference line and centerlines correctly overlay the +. Transfer the pattern to your plywood.
3 - now flip your paper pattern over to the right side of your intersecting lines.
4 - again align the paper 1/2 pattern so that the set level reference line matches your horizontal line but ignore the centerline alignment for now other than it should meet at the center of your + lines. Put a pin or nail in the center of the + to keep your paper pattern correctly placed.
5 - now get your your plans and find the detail of the frame you are laying out.
6 - check its top width dimension against the shape on the plywood, if it matches then great but if not then shift the right side paper pattern by pivoting on the pin - it should not need too much adjusting if any.
7 - now check the other frame dimensions on the plans against your work. They will be close enough to build your boat but you can fudge your full size pattern in step 10 below if you want. I would not. It will be good enough.
8 - transfer the correctly aligned right side pattern lines to your plywood.
9 - and last, draw a new correct centerline on your full size completed drawn pattern on the plywood. If there was an error in your centerline pattern in step 3 it should have been corrected or canceled out when you flipped the pattern over to do the right side. It makes no difference now since you have a correctly sized and drawn FULL SIZE frame pattern.
10 - now you have a correctly drawn and sized FULL size pattern to scan in if you want but I would just keep your original 1/2 pattern sheet as backup for next time and use the full size work to build the boat.

Paper patterns are just patterns. They are an aid to building. You could lay out the frames from the details on the drawings too and NEVER use the supplied Glen-L pattern sheet. The steps I outlined above correct any errors in much the same way that lofting does. You are expecting way too much from a sheet of blueline patterns if you think they will scan into Autocad and be correct to 2 decimal points! If you can build the boat to within 1/8th inch tolerance you are doing fine.

The photo below shows laying out my transom.

http://www.hevanet.com/kg7pv/Utility/image4.jpg

[Steve Miller]

Here is my first photo with the correctly highlighted centerline (blue) and setup reference level line (yellow). The red chaulkline shows the fore and aft orientation of the centerline.

http://www.hevanet.com/kg7pv/Utility/centerline.jpg

[bikemaniac]

Some of your answers have been difficult to understand so with all respect I have made a little picture and added some more text. Hopefully all of you can help me out.

I have tried to illustrate the problem by making a picture of the pattern:

Go to http://photos.yahoo.com/faksgnisner/ and click on "boatbuilding" and pick the 2nd picture called "glenl". You can download it for better viewing.

The left hand side illustrates the actual pattern very much exaggerated to simplify the problem. Note the angle which is less than 90 degrees - this shows the actual problem. The right hand side shows how the pattern really should be. To my opinion there are two solutions to my problem:

Solution 1:
Attach the frame to the centerline by means of the 2 fixpoints already touching both the frame and centerline (the uppermost and lowermost fixpoints). Then rotate the set up level ref line downwards in order to obtain a full 90,00 degrees angle (rotating around the intersection between centerline and set up level ref line). By doing that the middle fixpoint will change. This looks like the most natural and intuitive way to solve the problem. After this correction the entire picture is tilted counterclockwise to look like the right hand picture.

Solution 2:
Attach the frame to the set up level ref line by means of the single fixpoint (the middle one) which is already touching the set up level ref line. Then rotate the centerline counterclockwise to obtain a full 90,00 degrees angle (rotating around the intersection between centerline and set up level ref line). By doing that the uppermost and lowermost fixpoints will change. After this correction the entire picture is tilted clockwise to look like the right hand picture.

Now, I'm not going to do any changes on the frame itself, I just need to know if I should go for solution no 1 or solution no 2 and why. I don't have enough experience to choose the right one - that's why I'm asking the experts here :) . I'm not interested in fixing my problem on the go when building the actual frame by trimming here and there - I need to have the corrections done on the paper.

Thank you in advance, Lucas Jensen

[htom]

Ahh. My guess is that all three fixed points are correct, and what is wrong is the point of intersection of the centerline and the level line. Put the centerline perpendicular to the world, and raise that point of intersection to level the level line.

[Steve Miller]

I wonder if you have actually tried to do what I suggested in my long step by step post on 6-18-06? I mean on paper not just reading it.

You will correct your perceived error and have a functional pattern as a result. Remember that I have already suggested to draw out your own patterns using the info on the plans. It will then be correct to whatever level accuracy YOU can draw. If the patterns do not work for you then do NOT use them. I am not sure how I could explain this any better or in more detail than I already have but here goes:

I would not do either of your choices. Both will distort the shape more since the relationship of the side of the frame shape to the set up level reference line is the critical one. The goal here is NOT to correct your percieved defective Glen-L pattern but to make a new correct one that makes you happy. I still maintain, as the others already have, that the plans are what matters for accuracy and that the patterns are just a building aid and that the real error here is you expecting cad accuracy from the patterns that were hand drawn and copied a bunch of times. ie Copies of copies of copies of copies and so on......

So if you need another explanation please understand that the method posted below or by me on 6-18-06 are a method to make a new full size (not 1/2) pattern you can use to build your boat. I have already pointed out that you will need to do this to build the boat ...... If you still do not understand it then have some local boat builder do it for you so you can get building. If you have to have a back up set of patterns then have the sheet copied and FORGET scanning and cad and all that. The Glen-L boats were drawn by hand so live with that level of accuracy.


1. Make a an exact copy of your Glen-L frame pattern so you have a left and a right side. Trace it out. Forget scanning and printing!

2. Draw a line on your work surface or new piece of paper to represent the set up level reference line. Mark a center "fix point" on that line.

3. Align each of your patterns (left and right) to the set up level reference line and the center fixpoint as you call it. Ignore the vertical centerline.

4. The bottom fix point and set up level reference line and its center fix point should be very close or right on. Get them the best you can. Pin or tape the sheets so they can not move.

5. Now take a batten and spring a new line for the top of your frame that lines up on both the left side and the right side. You should be able to get a nice fair curve that matches your patterns pretty darn closely.

6. Last - draw a new correctly oriented vertical centerline.

7. You have a completed pattern that fairly represents the boat as drawn by Glen-L and is as accurate as needed to build from.

8. Now throw away your Glen-L patterns and use your new corrected one. The bottom line is you are over thinking this by a lot.

Good luck.