Scanning a steel tube aircraft fuselage?

Support
1

Greetings,

First post, and completely new to 3D scanning. I’ve watched these things for years, and recently bought a Vega scanner. I’ve played with it some to figure out what it can and can’t do, and I’ve had mixed success. It’s really counterintuitive in a way. You’d think something simple like a tube or a sphere would be the easiest thing, but in fact it’s just the opposite. Scanners seem to live for textures and features.

One of my rationalizations for buying this was to hopefully scan a small aircraft fuselage to get the dimensions for a drawing. I have a small engine mount made with similar construction, and I figured that would be an easy test to see if this was possible. So far, I haven’t been able to get a good scan of the mount since it just doesn’t seem to give the scanner enough textures or features. I’ve tried various scanning sprays and markers, but they didn’t help all that much. If you set the mount on a moving blanket, the scanner is much happier because it sees all those zig zag seams in the blanket. Unfortunately, it seems to be mostly scanning the blanket and the mount is just in the way. I’ve read about global markers, and also that it’s coming in the next software version. That might be my last hope.

I’ll try to post a few example pictures if I can figure out how. I’d be interested to know if this is just a hopeless job for the scanner, or if there’s just something I’m missing.

Thanks,
Rusty

engine mount 2
engine mount 21024×768 158 KB

2

Example fuselage.

Fuselage-1
Fuselage-11024×768 231 KB

3

I was hoping someone would have some suggestions for scanning these tube structures.

Does anyone know when the Vega will get the global marker feature?

Thanks,
Rusty

4

Hi,
It is a challenging objects, currently we suggest you to stick 12mm markers on the frame and keep a medium resolution in Fast Mode to scan and get the whole structure, if you need more details in some parts, you can use HD mode and 3mm markers to scan the part indoors and align the projects together. For this kind of slightly reflective metal, using spray can improve the data quality.

For global marker function, we will update it in the next one or two version, our team are working hard on it, we understand the importance to have it and the convenience it brings to.

5

Thanks for the comments. I haven’t tried 12mm markers, only 6 and 3, but for the small motor mount I’d hope 6 would be fine in fast mode. I’ll certainly try 12mm though. I really need to find some markers with repositionable tape that’s easier to remove. Probably can’t have too many markers.

6

Hi,
Maybe you can find some magnetic markers for this case, both 12mm marker and magnetic marker will save your preparation time.

7

I bought 1mm thick round magnets the size of the 3mm and 6mm markers, and they’re certainly faster and reusable. The fact that they’re flat on a round tube makes them subject to being moved, so you have to be careful. Also being flat, I’m not sure they provide as much orientation assistance as a curved sticker when you start moving around the tube.

I also tried some thin magnetic sheets cut to the size of the round markers. Pre-curving those would probably be good, though the magnetic force is so low that they just don’t want to stay in place.

On the engine mount, I used a normal black magic marker to make random markings in hopes that the scanner would used those, but it wasn’t any better really.

I even tried printing randomly generated patterns, and taping those around the tubes. That might have worked a bit better, though still not great. It’s certainly not practical for a full size fuselage.

Another test was to wrap a fairly thick string in a varying spiral around the tubes. I figured that would give it more to focus on, but there wasn’t any real magic there. I do think that method may have some promise though.

Yes, I’ve wasted a LOT of time tinkering with this :slight_smile:

Rusty

8

Hi,
For projects involving aircraft exterior structures, we generally recommend using metrology-grade 3D scanning equipment such as the FreeScan Nova Wireless Metrology 3D scanner. Entry-level scanners would require considerable effort for such applications.

Your proposed scanning accessories demonstrate impressive learning capabilities and hands-on initiative. We truly admire your dedication to engineering and 3D scanning technology. With professional responsibility, we must advise that this project requires industrial-grade scanning equipment. Even experienced scanning specialists might find this project challenging, and it could prove particularly difficult for someone new to 3D scanning technologies.

Should you require assistance, please feel free to visit shining3d.com to submit your contact information. Our technical experts would be pleased to provide professional consultation and support for your project. https://www.shining3d.com/

9

Thanks very much for your comments! It’s been interesting and (mostly) fun trying to figure out what the scanner can and can’t do, and I’m sure I’ll find more realistic uses for it.

I knew going into this that it was probably unrealistic to expect to be able to scan an airframe. I can manually measure the tube cluster locations to redraw the tube structure in CAD, but I will at least give the scanner a try once my airframe arrives. While it probably won’t scan the whole airframe, it still might be useful for localized areas where there are more complex details.

Since this is just a personal project, I could never justify the cost of professional tools.

Thanks again,
Rusty

10

I would try some aesub blue first on a small area and see if you can scan that section. It may be best to do smaller areas and mesh it in the software. If you find it does work you could try then going to the aesub orange and rescanning the whole craft as a secondary and mesh it with the sectional meshes. The blue will give you 4 hrs of time before it vanishes and orange will give you 12hrs.

11

I’ve tried the DIY 90% alcohol and cornstarch spray, as well as the commonly recommended Dr. Scholl’s foot spray. Both seemed to coat fairly well, but I didn’t get much success even scanning one corner of it. I do have a can of the Aesub blue, but didn’t figure it would be much different from the other sprays. The Aesub spray is crazy expensive, and I might have to take out a loan to buy enough for a whole fuselage :slight_smile:

For the most part, I just need to be able to locate the center of each tubing cluster, since it would all have to be redrawn in CAD anyway. I wouldn’t need much detail for the tubing between the clusters as long as it could keep track of where the clusters were located. A fuselage like this would be almost completely symmetrical too, so if I could just scan one side, that would probably be a good result.

I’ll more than likely end up using a plumb bob and paper taped to the floor to measure the fuselage. The scanner with $pray may still be useful for capturing details around attachment brackets.

Thanks,
Rusty

12

I wonder if you used texture mode and wrapped the tubing with colored electrical/painters tape if you could capture a section.
Also you are 100% right on the price of aesub blue. Have to sell a kidney. I have come across the aesub green for bulk items but your issue seems to be a geometry and tracking. Another thought you could possibly try is just put tape in small sections creating flags. This way you have markers. Might help the scanner not get lost. Good luck!

13

Thanks for the comments. I’ve tried erratically wrapping the tubes with thick colored string to give it more texture and color, but it wasn’t any huge improvement. I even printed random patterns and taped those around the tubes with no success. I see a new update just got installed, but the notes didn’t say anything about global markers. I do need to give this another try I suppose.

I might try 3D printing some odd shaped collars that would snap onto the tubes. That might give it more to orient to. Since I know the tube diameters, I don’t really care about the scan for tube details. What I need is the spatial location of the tubing joints. I may have even more reason to want to do this soon, but I’m not confident the scanner is going to do what I need.

14

I have been curious if the new Rigil would work any better for a tube frame? I doubt I could justify it.

15

Your project sounds very challenging. I do not know if my experience will help but I tired to scan a handlebar (carbon fibre part, black and matt surface) from my bicycle. Type of part is similar. A pipe without any featrue where the scanner can oriented. I tried:

  1. Without nothing
  2. With horrible expensive spray from AESUB
  3. With 3mm marker
  4. With 3mm marker and spray

Nothing worked out. This was done outside on a sunny day in shadow. I was thinking to sell my Vega as used scanner… I tried as last option in house at the evening with a lot 3mm marker like the hell. I just wanted get confirmation that it will not work and I have to sell the scanner. But surprisingly it worked! Ok, your part is bigger and such a number of marker on a long part does not work but if you want to scan a small area like reinforcement/stiffener area maybe it will work. Just my 2 cents.

DSC_2153
DSC_21531920×1440 193 KB

16

DSC_2152
DSC_21521920×1440 93.3 KB

17

I’ve been there as well considering selling the Vega. I’ve had some limited success in small areas using the HD mode. Oddly, I’ve been able to make it work just about as well with no prep as I have with spray or markers. I’ve had no real success with fast mode, and I’d need that to cover a larger area. Ideally, I’d use fast mode for the whole structure, then come back with HD for details like attachments for landing gear, lift struts, etc.

I recently tried fast mode again by 3D printing snap on collars that had irregular shapes on them. I snapped those around in different orientations thinking the scanner would be happy with that. It wasn’t. I also made some thin snap on collars for 12mm markers as earlier suggested, but those were no help either.

I’m thinking that HD mode can more easily cope with this because of the short focal range. It just doesn’t see anything else in the background. For fast mode, it wants to look at everything in the background as well. If I hang the engine mount from the ceiling, where there’s nothing close to it, the scanner doesn’t seem to see enough features to make it happy. If I set it on top of a moving blanket on the floor, it scans easily, BUT the scanner is focusing on the blanket, and the mount is just in the way with little detail. I’ve joked that fast mode will only scan things I don’t want. I can literally pick anything in the garage (belt sander, box of misc parts, etc) and it scans those just fine. I think it’s just messing with me :slight_smile:

I do appreciate the reply, and I’m glad you’re finding some success. I do think I can use this for smaller items, and details of brackets and such. Honestly though, for the relatively simple shapes I’ll be dealing with, I can do it faster by manually measuring and sketching the parts. I’m too stubborn to give up though.

Cheers,

Rusty

18

HI,

I too am new to this Vega Scanner however I have started applying some basics that are pointed out if you read up on scanners in general.

Firstly, scanners tend not to like white surfaces, anything black, or chromed. The chrome reflects the black at the reflective concentration point down the center of a rounded edge feature. This “concentration point” moves as you move since it is relative to the light and the perspective view.

I have hand some success scanning chrome that is slightly dirty or dull by moving around the chrome. This changes the reflective concentration points. When mirrored surfaces are bent the light they are reflecting becomes distorted. The light is concentrated in some places and diffused in others.

If you are scanning a tube remember that it is a bent surface and the light will concentrate down the center of the tube and if the tube is straight that offers the scanner no texture or features to pick up. This makes a tube one of the hardest things for a scanner to pick up ( a shiny sphere would be worse since the concentration point is a literal point - a tangent point on the surface)

I would suggest this. Stay away from black, white shiny surfaces. This scanner has a color mode. Might try using that for feature detection.

I would take masking tape (maybe green or yellow) and strap it across the various tubes to almost create surfaces between the tubes. The scanner needs features that it can reference. Make sure the tape is not shiny. You may even fit paper or cardboard in the spaces between the tubes.

If you use marking pens to draw shapes, do not use black ink.

Space the straps of tape so that at least three references can be detected by the scanner at any point along the scan.

Scanning surfaces with some changing form can be done with out marker dots. Bridging low feature ares is the challenge. I put random feature items along side of the subject so that the scanner can find a reference feature for tracking. A trash can with random stuff sticking up has worked.

I just did this today while scanning a front bumper to a sports car. The ends had enough features for tracking but the center was too simple. I put a trash can with rubbish (already provided) along one side and a bucket with a handle on the other side. This allowed me to cross the center and get the scan.

You might also consider placing colorful surfaces near the tubes so that they reflect the colors and not black.

Just my thoughts. Tubes are hard to scan.

~Terry

19

Thanks for the extensive comments Terry. I’ve tried nearly everything you suggested, but stringing tape between the tubes has some real possibilities.

I had concluded that the scanner was being confused by the large empty spaces between the tubes. If there’s nothing close, it just hasn’t got enough features to work with. If there is something close, like when I set that engine mount on a moving blanket, it’s far more interested in scanning the blanket than the tubes.

Some randomly placed tape between tubes is easy to do, and won’t have any effect on using the scans. The best I can hope for is finding the location of the tube joints (clusters) to enter into CAD so I’d just ignore the tubes and tape between clusters.

Funny thing is that I just got the fuselage I’d want to scan, and I’ve been cleaning years of dirt and spider webs off of it back to the shiny black powder coating. In retrospect, I should have tried scanning it before cleaning off all that excellent coating :slight_smile:

I’ll try the tape trick this week and see how that works on the engine mount.

Thanks,

Rusty

20

Thinking more about the tape, non adhesive tape with as much writing as possible on it might be a good option. Something like this maybe: 4Pcs Halloween Decorations Caution Fright Tape Bundle, Scary Caution Roll 3inx80ft Hazard Warning Tape Halloween Party Decorations Indoor Outdoor Spooky DIY Decorative Accessory (4PCS Tape Bundle): Amazon.com: Industrial & Scientific