Attention manufacturers

[sharkmouth]

Again, this is from my experience. Mumpeter (as Jennings called it) seemed to have the same problem Knitty Schmalk has, someone who resists making changes to CAD after the first renderings. In the case which I thought this thread was about, it was to have a company come up with a subject which was not on their to do list.

Mumpeter has responded very well to suggestions backed up with photographs and measurements (and I believe the same can be said for Knitty Schmalk as the only feedback from their F-101A is favorable). Back in 2012, I went to Poland and came back with several Gigabytes of digital photographs which included metric measurements of the items that interested me in the various museums. So far, the modern Soviet artillery and armor from them is getting very good feedback.

So, does anyone have the necessary references on the Sukhoi-15 so that the first CAD renderings would be correct? If so, then contact any of the companies you prefer and offer it to them. Do not forget to include a list of users and schemes which can be used to produce various releases which may differ in tiny details and/or markings. In your references, make sure to properly identify the variant in the photo (either written on the back of the print or in the file name) so variants do not get mixed up.

At the moment, I still have some companies asking me for armor suggestions based on the success of my past correspondence.

Regards,

[Jennings]

It's not even so much that they won't make changes in the CAD, it's that there is often little or no original research done in the first place. And the people doing the CAD design work clearly know little to nothing about real airplanes, and they appear unable to look at photographs and translate what they see into their CAD design. Everyone thinks CAD is this precise engineering tool - which it is. But in translating a real airplane into a scale model there is also a lot of artistic skill required to get it right. Being a talented CAD designer only gets you so far. If you don't have detailed knowledge of the subject, and of how real airplanes are constructed and how their various parts work and interact with one another, you'll never produce a really outstanding model. And cranking stuff out at the speed that KH does, there is no *way* you're going to get a consistently high quality product, since it's all coming from one designer.

[B.Sin]

Maybe there working Mr.Song to hard! LOL

[sharkmouth]

It's not even so much that they won't make changes in the CAD, it's that there is often little or no original research done in the first place. And the people doing the CAD design work clearly know little to nothing about real airplanes, and they appear unable to look at photographs and translate what they see into their CAD design.

Jennings, I agree but my experience shows me that the company (or it could be the designer) won't make the changes after the initial CAD renderings were completed. In my example, I approached Dragon with the Cadillac Gage Commando idea. Unlike a simple wish list suggestion which they won't heed, I explained the references I had which included photographs and measurements plus field and technical manuals plus marking schemes from 16 different countries. My pitch was that it fit several series (Vietnam, Modern, and Police). They listened, replied, and let it die.

Trumpeter accepted the help and thanked me. Then nothing, no feedback at all after acknowledging receipt of the materials. Why the silence? They used my R & D to create the series for Hobby Boss, their sister company. Not having any ties or contacts with Hobby Boss, I did not see any CAD renderings so I couldn't catch all the errors misinterpreted from photographs. While not a bad kit, it is nowhere what it could have been. It was a lesson learned.

If anyone has a photograph of an Su-15 with a shark mouth scheme, I may provide contact details for several companies to whomever has the references required to create a good kit.

Regards,

[Laurent]

Flow chart to try making an accurate model for small companies :)

[sharkmouth]

Laurent,

I like it but you left out the most important part... the 'push' behind the designers. If the company's owners only want to sell product, they may not allow the designer any time or funds (as in getting paid) to do revisions. This is why I think competition makes it worse (although some thought it should make it better). When there is competition (two companies announcing the same product) a race ensues and accuracy be damned. If there is no competition, time can be taken to release a kit with revisions taken into account.

Also, While CAD rendering can be drawn from plans, it is best to offer photographs and measurements along with said plans. Even if plans lack cross sections, oblique photographs make shapes clearer to see. For example, on road wheels, I took several perpendicular shots to show tread detail, tire detail, hub detail, a note written about flat spots on the bottom (museum item is empty and refer to in service photos), plus some oblique images showing the shape of the hubs.

Regards,

[Jennings]

You also left out "CAD designer has at least some knowledge of how real airplanes (or tanks or ships or spacecraft, or whatever) are designed, function, and how they are used in the real world."

Without that you could teach a chimpanzee how to do it.

[Freedom Models Martin]

I like the SU-11 whiles we're talking Soviet jets :P

[Laurent]

You also left out "CAD designer has at least some knowledge of how real airplanes (or tanks or ships or spacecraft, or whatever) are designed, function, and how they are used in the real world."

No. It just makes communication easier. From this post.

(Since the CAD drawers are no aircraft experts or even knows little about the aircrafts, you can imagine how tiring for us to explain everything to them. And to be honest, it's already a miracle for them to correct to this level I personally think)

[Laurent]

I like it but you left out the most important part... the 'push' behind the designers.

It isn't part of the process.

Also, While CAD rendering can be drawn from plans, it is best to offer photographs and measurements along with said plans.

Sure but not everything is convenient to measure. However some measurements can often be done to validate and "calibrate" the drawings.

[Jennings]

And let us not forget that CAD designers are not designing miniature airplanes. They are designing miniature sculptures (artistic representations) of airplanes. There is an element of artistic creativity required. Undefinable, but definitely required.

[B.Sin]

I guess for now I will have to struggle with the trumpeter kit ! I have the skills, its just a real pain in the you know what, and very time consuming.

Brad

[dehowie]

It isn't part of the process.

Sure but not everything is convenient to measure. However some measurements can often be done to validate and "calibrate" the drawings.

Well Eduard measured 4 different 109's and got the wingspan wrong buy about 15 inches and the fuselage the wrong diameter on the 109G-6.

How?

Oh plus out of scale gun pods..incorrect rudder..the list goes on.

So slight shape issues and perception problem are small compared to what war crimes Eduard committed to the 109G-6.

[B.Sin]

I like the SU-11 whiles we're talking Soviet jets :P/>

Hello Mr.Martin,

I believe Trumpeter is do an Su-11, so why do you guy's at Freedom Models be

a sport, and do a nice Su-15 for us.

Brad

[sharkmouth]

I like it but you left out the most important part... the 'push' behind the designers.

It isn't part of the process.

In an idyllic World, you'd be right but this is not that World. The push behind these designers are the investors and owners which, without them, the designers wouldn't even have a desk to sit at.

Also, While CAD rendering can be drawn from plans, it is best to offer photographs and measurements along with said plans.

Sure but not everything is convenient to measure. However some measurements can often be done to validate and "calibrate" the drawings.

That should be obvious for something as large as an airframe; an enthusiast would have a hard time providing that information with a tape measure. I gave a real World example of something I am involved in and having accurate measurements of certain parts makes it easier to rectify the visual impact of the larger items (perspective and proportion - the artistic elements Jennings alluded to).

Regards,

[ya-gabor]

Well Eduard measured 4 different 109's and got the wingspan wrong buy about 15 inches and the fuselage the wrong diameter on the 109G-6.

How?

Oh plus out of scale gun pods..incorrect rudder..the list goes on.

So slight shape issues and perception problem are small compared to what war crimes Eduard committed to the 109G-6.

Based on the interview with Stanislav Archman in the Eduard blog they measured only details and not general dimensions and used existing drawings in design.

"""We measure cuts, angles and partial dimensions that we connect with various drawings that are available."""

Although in the next sentence he said something else, so it is not completely clear if they made full measurements:

"""With recent experience, we are even forced to double check the overall dimensions."""

Best regards

Gabor

[punder]

Based on the interview with Stanislav Archman in the Eduard blog they measured only details and not general dimensions and used existing drawings in design.

"""We measure cuts, angles and partial dimensions that we connect with various drawings that are available."""

Although in the next sentence he said something else, so it is not completely clear if they made full measurements:

"""With recent experience, we are even forced to double check the overall dimensions."""

Best regards

Gabor

There is a solution for this problem: laser scanning. Su-15s are accessible in Slovakia and Lithuania. Scanning is expensive up-front but so is jacking around with bad drawings and selling kits that get panned for stupid-level accuracy.

[Jennings]

Laser scanning a 1/1 aircraft is *not* the be-all and end-all solution by any means. I know one specific instance very recently (the kit is still under design) where it resulted in major shape/proportion errors on the order of a scale foot or more in the size of a flying surface. Much depends on how the scanning is done and the skill of those doing the scanning.

How Eduard managed to screw things up as badly as they did using well known, long established published dimensions remains beyond my comprehension.

[punder]

Laser scanning a 1/1 aircraft is *not* the be-all and end-all solution by any means. I know one specific instance very recently (the kit is still under design) where it resulted in major shape/proportion errors on the order of a scale foot or more in the size of a flying surface. Much depends on how the scanning is done and the skill of those doing the scanning.

How Eduard managed to screw things up as badly as they did using well known, long established published dimensions remains beyond my comprehension.

Well yeah, they need to know what they're doing. :rolleyes:/>

Highly reflective surfaces can cause problems, along with inadequate training and User Error. But if done correctly, much less is left to the CAD guy for "artistic interpretation."

But you are correct... it is not a be-all/end-all. Just a very effective tool that imho is so far under-utilized.

I'd be interested to learn more about your example, but I understand if you don't feel free to discuss it.

Edited May 23, 2014 by punder

[B.Sin]

Whatever subject Mr.Jennings is talk about, I hope its shoots fire out of the back end,

and has red stars on wings.

Brad

[Laurent]

I don't know how 3D scanners are used but I guess that using a 3D scanner over a whole aircraft to obtain a digital object is very time consuming. It's not the scanning process itself that is but rather the merging of the overlapping surface patchs. The measurement error can be small for a single capture but the bigger the aircraft, the more captures need to be made and the more errors are introduced in the merging process I believe. And then after that, in any case, a clean object must be made out of the merged captures. It appears to me very time consuming and it wouldn't be surprising if the company preferred just taking the best drawings available (these perhaps ?) with additional inputs (no wing cross-sections so some additional reference material is needed for the cambered outer wing leading edge for example).

Edited May 23, 2014 by Laurent

[pookie]

As someone who worked with 3D computer design for a few years, I can say that creating an accurate rendition of an object of changing proportions is extremely difficult. In order to be as accurate as possible, the designer would need cross sections that are in the order of 2 to 3 inches apart (real world inches). That way you sketch each cross section individually then align them along a common axis and ask the program to extrude the shape.

I have yet to see a set of Aircraft drawings with cross sections every 2 to 3 inches. So what you have to do, is extrapolate the detail the cross sections miss and there in lies the problem. It's the extrapolation (filling in the blanks) of the details that some manufacturers miss and you end up with kits that have lots of errors.

When you work on a shape long enough, you develop tunnel vision. Some details get more attention while others do not. For every detail you get right, there are a few you get wrong, and those are the ones everybody else sees.

Edited May 24, 2014 by pookie

[punder]

As someone who worked with 3D computer design for a few years, I can say that creating an accurate rendition of an object of changing proportions is extremely difficult. In order to be as accurate as possible, the designer would need cross sections that are in the order of 2 to 3 inches apart (real world inches). That way you sketch each cross section individually then align them along a common axis and ask the program to extrude the shape.

I have yet to see a set of Aircraft drawings with cross sections every 2 to 3 inches. So what you have to do, is extrapolate the detail the cross sections miss and there in lies the problem. It's the extrapolation (filling in the blanks) of the details that some manufacturers miss and you end up with kits that have lots of errors.

When you work on a shape long enough, you develop tunnel vision. Some details get more attention while others do not. For every detail you get right, there are a few you get wrong, and those are the ones everybody else sees.

I drive Solidworks for a living and have tried in my spare time to build a 3D Su-15 from drawings. ALL the available drawings suck in some way, and the cross sections are always the least reliable part of the drawing. I suspect they are usually just "guesstrapolations" by the artist. Some of my better results have come from lofting between two widely separated cross sections, completely skipping the intermediates. But I don't have a model that satisfies me yet.

A few years ago I helped a guy scan a Cessna 310 at the Colorado Springs airport. It took a few hours as I recall. Still have the point cloud files. I will try converting them to surfaces at some, ah, point. But there are companies that scan big stuff for a living (usually buildings, industrial piping and such) and they already have the software, computing power, and expertise to do the job.

[Jennings]

Unless you had access to every single parts production drawing for the real airplane, and most likely unfettered access to the actual airplane itself, any "scale" drawing (as modelers understand the concept) is an artistic representation of the actual aircraft, and thus never, ever 100% faithful to the actual shapes. It's simply not possible.

[punder]

Unless you had access to every single parts production drawing for the real airplane, and most likely unfettered access to the actual airplane itself, any "scale" drawing (as modelers understand the concept) is an artistic representation of the actual aircraft, and thus never, ever 100% faithful to the actual shapes. It's simply not possible.

Exactly, that's my whole point. The technology is available for determining the actual shape of the airframe, without reliance on anyone's previous interpretation of that shape.