Free Internet Rowing Model (FIRM)

[NoEyeDeer]

Ugh. Coincidentally I just got a Delftship update to 7.xxx.

It's horrible. Interface has been completely upended, and graphics quality has gone down the tubes. Ok, so much for keeping it updated.

 

[Leo Lazauskas]

Ugh. Coincidentally I just got a Delftship update to 7.xxx.

It's horrible. Interface has been completely upended, and graphics quality has gone down the tubes. Ok, so much for keeping it updated.

Thanks for the warning. I just downloaded the update but haven't upgraded my version.
I only use it to generate Michlet offsets, so as long as that feature is still there I'll be happy.

 

[DCockey]

Ok, this is what I have at the moment. The file for the Herreshoff/Gardiner is done partially from memory (I used to see them every day), and partially from dimensions and drawings found online. The problem is with the latter, which everyone seems to have slightly different. Hey ho.

If someone has the actual table of offsets for any versions of this boat (original Herreshoff, or modified by Gardiner) I'd be happy to model it, on the basis of a gentlemen's agreement that I wouldn't use said table of offsets to build a boat. I'm not actually interested in purchasing plans for the boat since I'll probably never build one anyway, and the close approximation is good enough for my interests.

Same applies to the files for the Grant "Virginia" guideboat. I made two of these: one trimmed level and one trimmed aft one degree (guideboats are often rowed trimmed by the stern). The hull is only modeled to a height of 200 mm above the baseline, which is probably adequate. Can take it higher if necessary.

I also have a file D Cockey exported from Rhino, for a Rushton Saranac Laker (another guideboat). I haven't included that because it needs fairing and I still haven't got around to doing this. I'll make an effort on it, since the Saranac Laker is the prototype for the modern guideboats produced by some builders, so is probably one people are likely to encounter outside the Adirondacks.

Anyway, the other three are attached. No rights reserved. Do wotcha like.

1903-1906 Saranac Laker[/url]

I'm going to whip up a full model for this one. The term "Saranac Laker" is a generic Rushton term which covers quite a range of guideboats over the years, with different lengths and beams. This seems like a good example to use.

My Rhino model of the Rushton Saranac Laker was based on the offsets from the Adirondack Museum plans, presumably the same which NoEyeDeer found. Mean deviation of the Rhino surface from the published offsets is 0.004 inches, maximum deviation is .011 inches. The surface looks fair to me, certainly good for any analysis purpose or building a new boat.

The Saranac Laker plans were published in Rushton and His Times in American Canoeing by Atwood Manley, pp182-3, and are/were also available as single sheet from the Adirondack Museum. The museum also has plans for several other guideboats.

The original L. Francis Herreshoff rowboat design originally appeared as a lines drawing in the October 1947 issue of Rudder magazine, and were later reprinted in The Common Sense of Yacht Design p 137. I don't know of any published offsets by Herreshoff, and there may never have been any offsets.

Plans for John Gardner's (not Gardiner) interpretation were originally published in National Fisherman and were reprinted in Chapter 1 of Gardner's Building Classic Small Craft Volume 2.

 

[NoEyeDeer]

Thanks for the warning. I just downloaded the update but haven't upgraded my version.
I only use it to generate Michlet offsets, so as long as that feature is still there I'll be happy.

Yes it will still do Michlet offsets, once you find the carefully hidden export menu.

I've fallen back to 6.27 for the moment. I was idly considering upgrading to the Pro version, but with the graphics quality in 7.1 getting worse instead of better there's no way I'd pay for the thing at the moment. They seriously need to work on that. IMO it's no good having fast rendering if the quality is rubbish (even on the highest setting).

 

[Leo Lazauskas]

My Rhino model of the Rushton Saranac Laker was based on the offsets from the Adirondack Museum plans, presumably the same which NoEyeDeer found. Mean deviation of the Rhino surface from the published offsets is 0.004 inches, maximum deviation is .011 inches. The surface looks fair to me, certainly good for any analysis purpose or building a new boat.

The Saranac Laker plans were published in Rushton and His Times in American Canoeing by Atwood Manley, pp182-3, and are/were also available as single sheet from the Adirondack Museum. The museum also has plans for several other guideboats.

The original L. Francis Herreshoff rowboat design originally appeared as a lines drawing in the October 1947 issue of Rudder magazine, and were later reprinted in The Common Sense of Yacht Design p 137. I don't know of any published offsets by Herreshoff, and there may never have been any offsets.

Plans for John Gardner's (not Gardiner) interpretation were originally published in National Fisherman and were reprinted in Chapter 1 of Gardner's Building Classic Small Craft Volume 2.

Thanks, David.
I'll think about the hull I'll use as an example for a while.
I'll also try to include Olympic adaptive boats at some stage, but without the
outriggers. I have quite a lot of on-water data for them.

 

[NoEyeDeer]

My Rhino model of the Rushton Saranac Laker was based on the offsets from the Adirondack Museum plans, presumably the same which NoEyeDeer found. Mean deviation of the Rhino surface from the published offsets is 0.004 inches, maximum deviation is .011 inches. The surface looks fair to me, certainly good for any analysis purpose or building a new boat.

It's probably fair enough in general, but is missing the stem and sternpost, as well as the sheer, and doesn't have the chine at the edges of the bottom board.

It's also impossible to work with in Delftship, should anyone want to. The problem is that the way Rhino has exported the control net makes it useless in Delftship. You simply cannot modify it slightly and expect it to remain fair. To get a useful baseline model for experimentation in Delfsthip, it would be far easier to just make a new model, with a clean and usable control net, to the same table of offsets.

 

[NoEyeDeer]

Thanks, David.
I'll think about the hull I'll use as an example for a while.

TBH, if you only want to include one fixed seater by default, I'd be inclined to use a guideboat. This is simply because they are the benchmark for speed in traditional fixed seaters, and are definitely faster boats than the Herreshoff/Gardner boat.

The Michlet modelling I did a while back indicated that the Rushton Saranac Laker (David's ex-Rhino model) had slightly better resistance characteristics than the Grant Virginia (my model, that you already have), so IMO that would be the one to use.*

Graph attached.

*ETA: Although I'll still make a full model that is usable in Delftship, for my own reference if nothing else.

Incidentally, the graph was done by taking the three separate Michlet output screenshots, then doing some cunning cropping, overlaying, and vertical scaling in Photoshop. It's accurate for comparative purposes.

 

[DCockey]

It's probably fair enough in general, but is missing the stem and sternpost, as well as the sheer, and doesn't have the chine at the edges of the bottom board.

It's also impossible to work with in Delftship, should anyone want to. The problem is that the way Rhino has exported the control net makes it useless in Delftship. You simply cannot modify it slightly and expect it to remain fair. To get a useful baseline model for experimentation in Delfsthip, it would be far easier to just make a new model, with a clean and usable control net, to the same table of offsets.

I was referring to the original Rhino model which includes the entire hull surface to the sheer including the bottom board and stem. The .fbm file I distributed was based on the Rhino surface and was truncated at a "waterline" just below the mid-ships sheer. It has a set of points along the edges of the bottom board and stems.

Would you like the entire Rhino model exported in .wrl format which can be opened in DelftShip? (.wrl files can't be attached to posts) Also a .csv file with the published offsets is attached, units are feet.

Let me know if there is anything else I can help with.

 

[NoEyeDeer]

Yeah I'd be interested to take a look at the .wrl. You should be able to attach it here if it's zipped.

The .csv offsets are just confusing to me. I prefer the screenshot of the book's offsets (which I already have) since I find that easy to read.

 

[DCockey]

Attached is a .txt version of the offsets point file (with extraneous points deleted). It can be imported directly into DelftShip as a coordinate file, and contains the station, buttock and waterline offsets as a set of xyz point coordinates. The stem profile points will need to be duplicated and offset by half the width of the stem face. The file may or may not be useful, and is intended to be used with the table of offsets.

Also attached is a zip version of the .wrl file from the Rhino surface. It should be able to be uncompressed and imported into DelfShip.

 

[NoEyeDeer]

Thanks for those. I'm about to crash for the night, so I'll take a look at them tomorrow.

 

[NoEyeDeer]

I had a look at those two files. The Rhino export (.wrl) isn't much use in Deftship, except as a reference overlay. The thing is that Delftship really does not like triangles in the mesh. It does much better with quadrilaterals. Triangles really mess up the fairing in Delftship, so it's impossible to tell what you are dealing with. OTOH, I get the impression that Rhino totally loves triangles everywhere, so the two programs work on a totally different basis.

This also applies to the density of the control mesh. I don't know how Rhino handles mesh manipulation, but getting a usable result in Delftship requires a much simpler mesh. If you throw in too many vertices it tends to get unmanageable.

Anyway, what I ended up doing was taking the .txt file of the basic coordinates and building a new Delftship model from that. The offsets given in the book are obviously not fair in places (the stem profile about 2/3 of the way up being a good example) and are missing some information, so I took them as guidelines and built a model that approximates the given offsets as closely as possible, while still having a fair surface, and working within Delftship's limitations.

It's looking pretty good now, but is still not quite finished. It turned out to be almost identical to your old .fbm export in terms of BMT and wetted surface, but with a slightly higher prismatic. The really interesting thing is the residuary resistance predictions from Michlet. These are lower for the new Delftship model, by a significant amount.

At 2 knots the new model has R_r of 0.000117 kN compared to the "Rushton GB 003M" with 0.000279 kN. So, at 2 knots the residuary resistance of the new model is 42% of the old one.

At a more useful speed of 5.47 knots, the new model has R_r of 0.012825 kN, compared to the older model's 0.014104 kN, meaning the model has just over 9% less residuary resistance.

Now obviously these don't equate to the same reduction in total resistance, but there's still a difference of 3-3.5% in total resistance over the range from 2 knots to 7 knots.

 

[DCockey]

The Rhino model was created using NURBS surfaces, not meshes. The NURBS surface for the "planking" has 9 x 13 control points. The mesh in the .wrl file was generated for export from the Rhino NURBS surfaces but is not directly related to the control point arrangement.

I'm curious about the differences between your DelftShip model and my Rhino model. Could you export your model as DXF mesh, Wavefront file (OBJ) or STL and post it?

 

[NoEyeDeer]

Yeah I'll post the current version of it, but I just did a silly thing.

I was messing around with it, trying to get it a bit fairer up near the forefoot, which can be a bit tricky in Delftship. Got it looking generally fairer there, but that's gone and increased the predicted resistance. This would be ok, except some silly bugger forgot to save the previous version with a different name. Result is that I now have a new version with more resistance, and will have to figure out how to reconstruct the old version again if I want a lower resistance version.

This is interesting though, because it's giving ideas for guideboat tweaking. I'll have to play around with a few of them as I get time.

Anyway, this is what I have at the moment. The zip has both .dxf and .obj files in it. I just zipped them because they're a lot smaller that way.

 

[NoEyeDeer]

Hey just for comparison, here's the two out.mlt files for both my versions: the good (but less fair) one I just lost and the current one.

Also the out.mlt file for your old GB_003M. These are all at 0.112 m³ immersed volume, which is about right for me and a boat. Speed range is 2 knots to 7 knots.