Reverse engineering sw

So we are ordering tomorrow the David SLS3 scanner, and we received this weekend our FDM printer (an advanced model still not on the market).

Generated the point cloud from the scanner we want to convert it into 3d model both for printing but also for other uses such as presentation or further CAD modifications.

What can be a good to start reverse engineering sw? Rhino does have his plugins but appear to be a steep start, MOI3D is not designed for that, was looking at Geomagic but this is simply not my area of knowledge so I can't evaluate it.

The David scanner will produce STL files, which can be printed directly. If there are errors in the files like small holes, overlapping triangles, etc. there are free or inexpensive applications (Netfabb, Meshlab, etc.) which will clean them up sufficiently for printing or presentation.

Reverse engineering (RE) is different, though. This is required for mechanical sorts of things where precision is important. With RE, you can identify a feature like a hole, slot or boss, make it a specific size, retrieve the "design intent" and incorporate it in a CAD model so a functional part can be recreated by 3D printing or CNC machining.

Geomagic currently has a couple of RE applications: Geomagic Wrap and Design X. Wrap is less expensive and easier to use; Design X is more powerful. We're resellers for both products, so let us know what exactly you're trying to achieve, and we can help decide what you need to get it done.

Andrew Werby
www.computersculpture.com

Quote:

---

Originally Posted by davide445

So we are ordering tomorrow the David SLS3 scanner, and we received this weekend our FDM printer (an advanced model still not on the market).

Generated the point cloud from the scanner we want to convert it into 3d model both for printing but also for other uses such as presentation or further CAD modifications.

What can be a good to start reverse engineering sw? Rhino does have his plugins but appear to be a steep start, MOI3D is not designed for that, was looking at Geomagic but this is simply not my area of knowledge so I can't evaluate it.

---

Quote:

---

Originally Posted by awerby

The David scanner will produce STL files, which can be printed directly. If there are errors in the files like small holes, overlapping triangles, etc. there are free or inexpensive applications (Netfabb, Meshlab, etc.) which will clean them up sufficiently for printing or presentation.

Reverse engineering (RE) is different, though. This is required for mechanical sorts of things where precision is important. With RE, you can identify a feature like a hole, slot or boss, make it a specific size, retrieve the "design intent" and incorporate it in a CAD model so a functional part can be recreated by 3D printing or CNC machining.

Geomagic currently has a couple of RE applications: Geomagic Wrap and Design X. Wrap is less expensive and easier to use; Design X is more powerful. We're resellers for both products, so let us know what exactly you're trying to achieve, and we can help decide what you need to get it done.

Andrew Werby
www.computersculpture.com

---

Wrap and Design are in the $Nx000 range as I understood, something more than we want to invest for a start, considering this is not the company primary business but a new line of services we need to develop both as leads and skills.
Interested to know options with pricing on par or below Rhino/Modo + plugins. Also interested to know the sw pipeline for different usage of the point cloud, solid modelling for CAD and printing and polygonal for animation and presentation.

Hi,
I was shown some reverse engineering software in 2009 : you had to manually select points and detail what you expected should be there (plane, cylinder, ...). From my point of view, given the status of the soft at the time it was faster to measure on the part and redesign it yourself. Added bonus of doing the design : it forces you to think your tolerances properly. There was also the issue that the result surfaces were close to the intent but not perfectly so which might create issues for assemblies as mating faces may end up slighly at an angle.

Quote:

---

Originally Posted by davide445

Wrap and Design are in the $Nx000 range as I understood, something more than we want to invest for a start, considering this is not the company primary business but a new line of services we need to develop both as leads and skills.
Interested to know options with pricing on par or below Rhino/Modo + plugins. Also interested to know the sw pipeline for different usage of the point cloud, solid modelling for CAD and printing and polygonal for animation and presentation.

---

Wrap is (in my opinion) the best software for turning your point cloud into a mesh and basic mesh alterations. The David software will also do this however, and to start with I agree it's an avoidable expense for reverse engineering applications. Wrap does an auto surfacing function which has its occasional uses.

Rhino with T-Splines add-on is a great affordable option for surfacing scan data.

For solid modelling we use Spaceclaim. Solidworks etc. will all work as well, but Spaceclaim hands itself nicely to Scan Data and in my experience is by far the quickest for producing results. In a standard CAD package you're purely using the scan data as a template. Cut planes through it, sketch around the curvature and extrude up to them. Spaceclaim will let you snap up to a mesh point, but you want to be working to design intent. Use the scan for the shape data and as a reference but use other metrology tools for taking critical dimensions. You have to be careful with programs like Design X that you don't make the height of your object 9.9374mm tall and all the planes off by 0.2 degrees.

With Spaceclaim there is a 3D scanning module (Geomagic capture) but it was over £5k and it's seldom we use it (though it does have lots of point cloud to mesh and mesh alteration features, which if you don't have wrap...) The 3D printing module can be very useful when working directly in meshes for merging meshes and using solid geometry for cutters (scanning something, filling the holes, and then cutting clean ones out before printing.) This is part of the standard spaceclaim license which could prove useful: https://www.youtube.com/watch?v=sesc4ZrZ9fo

Quote:

---

Originally Posted by JSenior

Wrap is (in my opinion) the best software for turning your point cloud into a mesh and basic mesh alterations. The David software will also do this however, and to start with I agree it's an avoidable expense for reverse engineering applications. Wrap does an auto surfacing function which has its occasional uses.

Rhino with T-Splines add-on is a great affordable option for surfacing scan data.

For solid modelling we use Spaceclaim. Solidworks etc. will all work as well, but Spaceclaim hands itself nicely to Scan Data and in my experience is by far the quickest for producing results. In a standard CAD package you're purely using the scan data as a template. Cut planes through it, sketch around the curvature and extrude up to them. Spaceclaim will let you snap up to a mesh point, but you want to be working to design intent. Use the scan for the shape data and as a reference but use other metrology tools for taking critical dimensions. You have to be careful with programs like Design X that you don't make the height of your object 9.9374mm tall and all the planes off by 0.2 degrees.

With Spaceclaim there is a 3D scanning module (Geomagic capture) but it was over £5k and it's seldom we use it (though it does have lots of point cloud to mesh and mesh alteration features, which if you don't have wrap...) The 3D printing module can be very useful when working directly in meshes for merging meshes and using solid geometry for cutters (scanning something, filling the holes, and then cutting clean ones out before printing.) This is part of the standard spaceclaim license which could prove useful: https://www.youtube.com/watch?v=sesc4ZrZ9fo

---

David sw will generate the mesh also? I did think he generate only the point cloud, even if with alignment.
About the others all are for sure wonderful sw but also expensive and difficult to learn.

We can say the cheapest (even if not the easier to use) option is Rhino+plugin?

What about MOI3D? I'm always returning to Moi since appear to be the easy start for sold modelling, less scary for beginners as us.

Just to understand the difference: tools such as SpaceClaim does have specific tools for reverse engineering, or they are just good solid modelling tools that fit nicely into the workflow that use the point cloud as a basis for modelling? What's the point where Wrap add value vs a generic solid modeller?

Quote:

---

Originally Posted by davide445

David sw will generate the mesh also? I did think he generate only the point cloud, even if with alignment.
About the others all are for sure wonderful sw but also expensive and difficult to learn.

We can say the cheapest (even if not the easier to use) option is Rhino+plugin?

What about MOI3D? I'm always returning to Moi since appear to be the easy start for sold modelling, less scary for beginners as us.

Just to understand the difference: tools such as SpaceClaim does have specific tools for reverse engineering, or they are just good solid modelling tools that fit nicely into the workflow that use the point cloud as a basis for modelling? What's the point where Wrap add value vs a generic solid modeller?

---

David will create a mesh yes (I'm 99% sure anyway - have never used it.) In some ways it's more accurate working from a point cloud but I nearly always convert to a mesh first.

It really depends what you're doing whether Rhino is a good tool. If you're primarily doing bodywork then great. It's one of the best until you have £30k to spend. It's slow for parametric modelling however.

Wrap is completely different to a solid modeller. It can (sometimes) create a closed (but ugly) surface from a watertight mesh, but this is a small feature of the program. With most hardware we can scan directly into the program, align the scans, merge them into meshes and then you have lots of tools to smooth/clean the data (which you wouldn't necessarily do for reverse engineering.) There are other programs that can do the most of what Geomagic Wrap can do - it is just good, simple and quick. If you're working meshes you really want to be learning Z-Brush as well.

Spaceclaim has a few scanning specific elements (and it handles meshes much better than most other mainstream programs,) but most auto processes I would recommend to stay away from. 3D scanning makes producing a 3D model of all but the simplest items quicker and more accurate, but you still have to do the modelling, so the best/quickest CAD program for yourself I would always recommend.

In the reverse engineering side of my company I would estimate I spend 5% my time in Wrap, 90% in Spaceclaim and perhaps 0.5% in Rhino with a few other programs added in.

I've no experience with MOI3D

Quote:

---

Originally Posted by JSenior

David will create a mesh yes (I'm 99% sure anyway - have never used it.) In some ways it's more accurate working from a point cloud but I nearly always convert to a mesh first.

It really depends what you're doing whether Rhino is a good tool. If you're primarily doing bodywork then great. It's one of the best until you have £30k to spend. It's slow for parametric modelling however.

Wrap is completely different to a solid modeller. It can (sometimes) create a closed (but ugly) surface from a watertight mesh, but this is a small feature of the program. With most hardware we can scan directly into the program, align the scans, merge them into meshes and then you have lots of tools to smooth/clean the data (which you wouldn't necessarily do for reverse engineering.) There are other programs that can do the most of what Geomagic Wrap can do - it is just good, simple and quick. If you're working meshes you really want to be learning Z-Brush as well.

Spaceclaim has a few scanning specific elements (and it handles meshes much better than most other mainstream programs,) but most auto processes I would recommend to stay away from. 3D scanning makes producing a 3D model of all but the simplest items quicker and more accurate, but you still have to do the modelling, so the best/quickest CAD program for yourself I would always recommend.

In the reverse engineering side of my company I would estimate I spend 5% my time in Wrap, 90% in Spaceclaim and perhaps 0.5% in Rhino with a few other programs added in.

I've no experience with MOI3D

---

We are not doing minime for sure...the first object we have request to scan is the propeller of an amphibious military vehicle.

Giving cleaned mesh to CAD staff is normally enough to enable them to reproduce the original design?
Or this approach does have little consulting value and we need to invest for a real solid modeling tool and skills?

Quote:

---

Originally Posted by davide445

We are not doing minime for sure...the first object we have request to scan is the propeller of an amphibious military vehicle.

Giving cleaned mesh to CAD staff is normally enough to enable them to reproduce the original design?
Or this approach does have little consulting value and we need to invest for a real solid modeling tool and skills?

---

Perhaps 10% of our work we're providing pure scan data (meshes or point clouds.) Most people are looking for an end result - whether it's an inspection report, finished solid model or the end manufactured product. Coming to you is just a way to get there so the more stages you can offer or at least assist with the (have a chain of people to recommend) the better.

You'll come across a lot of people who have had bad experiences with scanning in the past as they don't understand the data (they were probably given what they asked for and not what they needed,) and the company doing it has run away without offering any help.

By all means offer pure scan data, there is a market for it, but aim to be clear with what you're offering. Once you can successfully model there will be a lot more work out there.

Quote:

---

Originally Posted by davide445

We are not doing minime for sure...the first object we have request to scan is the propeller of an amphibious military vehicle.

Giving cleaned mesh to CAD staff is normally enough to enable them to reproduce the original design?
Or this approach does have little consulting value and we need to invest for a real solid modeling tool and skills?

---

Really informative thanks.

Returning into solid modelling I did discovered DesignSpark Mechanical, appear to be the stripped down free version of SpaceClaim. Any reason to don't start with that? I read it's integrated into RS Components assets DB, so it's more focused on PCB or can be used also for something like our propeller?

Need to say, one reason I'm so reluctant in starting with a CAD program is the horrible UI they got, appear all are just remained into the '90. No no other modern/easy to start with solid modelling sw?

Designspark Mechanical is great for making 3D printing models but you can't save out as a solid model (though somebody with Spaceclaim can do this for you.) We have an apprentice here just starting on it for the project for his course.

It's a bit stripped down but most of the features are there.

The reason I like Spaceclaim for 3D scan data (I'm not a dealer) is that it is very visual. It's very different to most CAD packages and a lot easier to learn.

Have a try and see what you think.

Quote:

---

Originally Posted by davide445

Giving cleaned mesh to CAD staff is normally enough to enable them to reproduce the original design?
Or this approach does have little consulting value and we need to invest for a real solid modeling tool and skills?

---

What do you intend to do with it ?

If it's just to make a replica for a static display, go ahead (though there are less expensive ways to go). If you intend it to be a usable component... don't put your hopes up too much. I had such a request for a turbine 1st stage : I refused as there was no way for me to ensure proper balancing, tolerances to the casing... For an external water propeller you won't have as much issues probably but you have to be reaslistic as to what this allows you to do.

Quote:

---

Originally Posted by JSenior

Designspark Mechanical is great for making 3D printing models but you can't save out as a solid model (though somebody with Spaceclaim can do this for you.) We have an apprentice here just starting on it for the project for his course.

It's a bit stripped down but most of the features are there.

The reason I like Spaceclaim for 3D scan data (I'm not a dealer) is that it is very visual. It's very different to most CAD packages and a lot easier to learn.

Have a try and see what you think.

---

Downloaded and appear to be nice. Tried importing some example of DAVID scans in obj format and is working fine. Now just need to understand how to use this scans as a guide for modelling.

Regarding the creation of mesh from the point cloud, confirm DAVID 4 sw can create it, not sure how advanced is into generating clean and correct one. What can be a nice and cheap program to correct original meshes? Meshmixer? MeshLab? 3D-Coat?

Quote:

---

Originally Posted by LambdaFF

What do you intend to do with it ?

If it's just to make a replica for a static display, go ahead (though there are less expensive ways to go). If you intend it to be a usable component... don't put your hopes up too much. I had such a request for a turbine 1st stage : I refused as there was no way for me to ensure proper balancing, tolerances to the casing... For an external water propeller you won't have as much issues probably but you have to be reaslistic as to what this allows you to do.

---

This is just a first request and we still need to receive the scanner, so we need to qualify the request itself.
Do you refuse to manufacture/print the turbine stage or to scan it?

Geomagic Design X is probably what you are after. Not to be confused with Geomagic Design, which is an MCAD tool.

I'd love to have Spaceclaim. Been using DesignSpark for a while now. Why do they have to make all these packages so damned expensive!

So our first work appear to be really the scan of a nautic propeller. Goal: reproduce the 3d CAD design since it's an old model and no 3d model is available, so to be able to produce it using metal printing.
Discovered the company (part of a bigger one) does have an CATIA license and workstation available.
Starting with no real modelling experience such as ours, CATIA REV tools will be really helping us or its a too steep path to start with? DesignSpark will be able to easier model a complex geometry such as a propeller just following the imported mesh?

Quote:

---

Originally Posted by davide445

So our first work appear to be really the scan of a nautic propeller. Goal: reproduce the 3d CAD design since it's an old model and no 3d model is available, so to be able to produce it using metal printing.
Discovered the company (part of a bigger one) does have an CATIA license and workstation available.
Starting with no real modelling experience such as ours, CATIA REV tools will be really helping us or its a too steep path to start with? DesignSpark will be able to easier model a complex geometry such as a propeller just following the imported mesh?

---

Would prolly be easier to just design it scratch i.e. by measuring it (the other reverse engineering).

Quote:

---

Originally Posted by davide445

So our first work appear to be really the scan of a nautic propeller. Goal: reproduce the 3d CAD design since it's an old model and no 3d model is available, so to be able to produce it using metal printing.
Discovered the company (part of a bigger one) does have an CATIA license and workstation available.
Starting with no real modelling experience such as ours, CATIA REV tools will be really helping us or its a too steep path to start with? DesignSpark will be able to easier model a complex geometry such as a propeller just following the imported mesh?

---

It really depends how simple the propeller is. Is it a single thickness? In the simplest method, you're effectively going to have to load in the scan and then model a 2D shape for the start and finish points (oversize the shape and orientation wants to be correct) and then draw guide curve splines from corner to corner. You then want to loft between them which I don't think you can do in DSPK (only Spaceclaim) - this should be simple in CATIA however, if you transfer the files after. You'll then want to draw around (in whichever program) the profile of the propeller, and cut this away from the solid you make. You can then model your boss, rotationally mirror the propeller around it, combine all and add in your rads as appropriate.

If it's not a single basic thickness this isn't necessarily the best way, but it depends on what they are wanting it for, and the tolerances required.


To Sebastian: 3D scanning would achieve far quicker and better results in this instance than any traditional method.

Quote:

---

Originally Posted by JSenior

It really depends how simple the propeller is. Is it a single thickness? In the simplest method, you're effectively going to have to load in the scan and then model a 2D shape for the start and finish points (oversize the shape and orientation wants to be correct) and then draw guide curve splines from corner to corner. You then want to loft between them which I don't think you can do in DSPK (only Spaceclaim) - this should be simple in CATIA however, if you transfer the files after. You'll then want to draw around (in whichever program) the profile of the propeller, and cut this away from the solid you make. You can then model your boss, rotationally mirror the propeller around it, combine all and add in your rads as appropriate.

If it's not a single basic thickness this isn't necessarily the best way, but it depends on what they are wanting it for, and the tolerances required.


To Sebastian: 3D scanning would achieve far quicker and better results in this instance than any traditional method.

---

We will try using CATIA next days with some test files (we need to still receive the scanner).
A bit scared to approach such a monster parametric solid modeller without any training nor experience in CAD modelling.
Simply I can't find a cheap, easy and RE powered solution in the $1000 range.
(Autodesk) Delcam PowerShape-e it's free with also RE features, but need to try how easy is to use.
Rhino + some plugins is around $2000
Can't find any info about the pricing of 3dreshaper Meteor, so maybe this is also an option.
VRMesh Reverse it's in the $1200 range but appear to be more oriented in architectural reverse modelling.
Also did find BricsCAD Pro and PunchCAD Shark LT less than $1000, again no idea about the usability, nor they have RE specific features AFAIK.

Ok so received the scanner we are testing it.
First work before the propeller one is a pair of eye wear.
Understood before thinking about RE we need to have a clean mesh to work with.
I can use DAVID 4 and other free sw to create and clean the mesh, but I still need to remodel where the scan is not complete on tiny or internal parts such as the hollows or guides.
I did read good things about 3d-coat, any other suggestion?

You don't need a perfect mesh to do RE; you just need enough of it to reconstruct important features. So, for instance, if you had a cylindrical boss on your original part, all you'd need was three good points on the mesh to recreate the circle, and one on top to show you where to end the extrusion. The mesh is just for reference, and once you've created the CAD model you can dump it - the CAD model is your final product.

3DCoat is intended for modeling organic objects, texturing them with UV maps, and remeshing them to clean up your models, but I don't think it makes any claims for being a reverse-engineering tool, or something that will perform mesh repair on scans.

Andrew Werby
www.computersculpture.com

Quote:

---

Originally Posted by awerby

You don't need a perfect mesh to do RE; you just need enough of it to reconstruct important features. So, for instance, if you had a cylindrical boss on your original part, all you'd need was three good points on the mesh to recreate the circle, and one on top to show you where to end the extrusion. The mesh is just for reference, and once you've created the CAD model you can dump it - the CAD model is your final product.

3DCoat is intended for modeling organic objects, texturing them with UV maps, and remeshing them to clean up your models, but I don't think it makes any claims for being a reverse-engineering tool, or something that will perform mesh repair on scans.

Andrew Werby
www.computersculpture.com

---

Difficult to understand the complete workflow.
In that perspective you suggest we need to use a CAD and remodel from scratch the object, using the mesh just as reference to ease the work.
I suppose this is just fine for mechanical parts with easy geometry, made from a mixture of solid primitives.
But with more complex objects such as eyewear or the propeller will be not better to have the full mesh, using specialized tools such as VRMesh Reverse, 3dreshaper Meteor and SpaceClaim to first reconstruct the complete mesh and next model a complete solid using semiautomatic generation tools that will use the complex mesh as guide?

Quote:

---

Originally Posted by davide445

Difficult to understand the complete workflow.
In that perspective you suggest we need to use a CAD and remodel from scratch the object, using the mesh just as reference to ease the work.
I suppose this is just fine for mechanical parts with easy geometry, made from a mixture of solid primitives.
But with more complex objects such as eyewear or the propeller will be not better to have the full mesh, using specialized tools such as VRMesh Reverse, 3dreshaper Meteor and SpaceClaim to first reconstruct the complete mesh and next model a complete solid using semiautomatic generation tools that will use the complex mesh as guide?

---

Have you got an image of the eyewear you can share?
For most reverse engineering applications (depending the reason for RE and the data format they require) you're using the mesh as your template/guide and cleaning the mesh serves little to no point. It will only make the data less accurate.

Quote:

---

Originally Posted by JSenior

Have you got an image of the eyewear you can share?
For most reverse engineering applications (depending the reason for RE and the data format they require) you're using the mesh as your template/guide and cleaning the mesh serves little to no point. It will only make the data less accurate.

---

Attached the first test scan, only of the frontal part without the temples

Quote:

---

Originally Posted by davide445

Attached the first test scan, only of the frontal part without the temples

---

For the basic shape you're going to want to create a surface over the front face and then draw around the profile of the glasses, project this on your surface and delete the excess. Once you have the basic curved shape the extrude it by x mm (assuming it is consistent) and can add some rads to the edges. If the lenses are recessed you can make a cutter using a section of what you've already created.

You can do this in Spaceclaim/Rhino/Most other solid modellers (It can be done in Designspark but creating the front face is difficult as there are no blend/surfacing tools.) There's no point trying in a mesh program however, as you won't get consistent sharp edges. Depending how clean your mesh is, you may want to smooth the front face to be able to create a smoother surface.

This will all take you time to learn however. When I purchased my first scanner (and I was using CAD before this) I spent 6 months doing no paid work before I was comfortable taking on clients projects.

Quote:

---

Originally Posted by JSenior

For the basic shape you're going to want to create a surface over the front face and then draw around the profile of the glasses, project this on your surface and delete the excess. Once you have the basic curved shape the extrude it by x mm (assuming it is consistent) and can add some rads to the edges. If the lenses are recessed you can make a cutter using a section of what you've already created.

You can do this in Spaceclaim/Rhino/Most other solid modellers (It can be done in Designspark but creating the front face is difficult as there are no blend/surfacing tools.) There's no point trying in a mesh program however, as you won't get consistent sharp edges. Depending how clean your mesh is, you may want to smooth the front face to be able to create a smoother surface.

This will all take you time to learn however. When I purchased my first scanner (and I was using CAD before this) I spent 6 months doing no paid work before I was comfortable taking on clients projects.

---

Useful suggestions, thanks. Using SpaceClaim we will have RE tools available, using "clean" Rhino without any RE addon such as RhinoResurf or Mesh2Surface will be difficult? Trying to understand the most cost/effective investment for our learning, considering we are just starting.
We will probably proceed step by step, starting with just the scan and mesh preparation, and giving to colleagues the difficult reverse part and next learning how to.
A second project is a shark tooth (attached the first scan), a definitely more organic shape. We need later to print it in 3d: we will need also to pass trough a CAD and solid modelling or we can stay with the mesh and just give them a depth and make it watertight?

Assuming they want a like for like model then no reason to put into CAD. Scan, mesh, fill the holes, smooth out any problems, align it to the world axis and then maybe cut a plane off the bottom and fill so it will sit flat.

Quote:

---

Originally Posted by JSenior

Assuming they want a like for like model then no reason to put into CAD. Scan, mesh, fill the holes, smooth out any problems, align it to the world axis and then maybe cut a plane off the bottom and fill so it will sit flat.

---

In this case any suggestion for a good tool?
I think I did now understand the CAD topic, about mesh there are many more options, I did have on my laptop for testing Meshmixer, Netfabb Private, 3d-Coat, Meshlab, VRMesh Reverse, GOM Inspect, Cinema 4D, will need a lot of time just testing every of them.

It looks like the first step is to resolve the two scans you got of that shark tooth into one. Have you been able to do that with the software supplied with your David scanner? I agree - you don't need reverse engineering to get from a STL file like that (once you've unified and cleaned it up to produce a water-tight mesh) to print.

Andrew Werby
www.computersculpture.com

Quote:

---

Originally Posted by awerby

It looks like the first step is to resolve the two scans you got of that shark tooth into one. Have you been able to do that with the software supplied with your David scanner? I agree - you don't need reverse engineering to get from a STL file like that (once you've unified and cleaned it up to produce a water-tight mesh) to print.

Andrew Werby
www.computersculpture.com

---

I was able to use DAVID sw to create the full STL model and print it.
Now back on my original task of RE the eyewear, tested SpaceClaim and a bit Rhino.
Pretty similar features (STL cutting and contours blending or just reskinning), did find difficult to understand in what kind of RE projects will be advantageous using a solid CAD such as SpaceClaim vs a surface oriented CAD such as Rhino.

Once the love affair with 3D scanners has subsided I bet you end up just modelling it from scratch(like SF suggested).

No, but seriously, you could model a propeller in maybe 1000 vertices.

Quote:

---

Originally Posted by Nihonddd

Once the love affair with 3D scanners has subsided I bet you end up just modelling it from scratch(like SF suggested).

No, but seriously, you could model a propeller in maybe 1000 vertices.

---

For the propeller might be the best idea, for the eyewear I'm now working on having a 3d mesh as a guide is really useful.
Every case will need to right tools.

So SpaceClaim it's really good but simply cost too much for the features.
Rhino does have the same features (mesh import, slicing, countours blending, resurfacing) but I didn't find myself comfortable with the whole UX.
Any other option in the same Rhino price range?

Quote:

---

Originally Posted by LambdaFF

Hi,
I was shown some reverse engineering software in 2009 : you had to manually select points and detail what you expected should be there (plane, cylinder, ...). From my point of view, given the status of the soft at the time it was faster to measure on the part and redesign it yourself. Added bonus of doing the design : it forces you to think your tolerances properly. There was also the issue that the result surfaces were close to the intent but not perfectly so which might create issues for assemblies as mating faces may end up slighly at an angle.

---

Agree with this

Create eyewear bow from zero would be long, the curve is changing differently in the upper and lower side, also the thickness is changing in different parts and from top to bottom.
What would you use to model this, for a price tag such as Rhino.