I fear I am in need of some therapy as modeling this thing has become a bit of an obsession. I have never looked so intently at a Ford in all my life! That speaks volumes to a degree you cannot understand. Half my family tree worked for GM and other half Ford. Christmas dinner was an interesting summit meeting!

Once body design become ingrained in one's mind, anything that is not perfect will not, can not, must not be let go until it's fixed. It simply ruins the ability to "let it go". As a result, I simply HAD to begin rebuilding the front fender. It wasn't correct and I knew my Sunday would not be peaceful until it was. Yes, I admit, it's a sickness. Jolyon, you said you were a car guy...maybe you can relate?

UGMENTALCASE, I have noticed the need to "watch" everything and make sure things are joined. I have done exactly your suggestion a few times as the data was not cooperating as desired. I did find success with my splines. Not necessarily staying attached to entities I attached them to, but I was able (with Rokoko's help) to go back and modify the spline attachments and their tangents (Shark really needs G2, desperately), Update Links and it all seemed to behave as expected.

Rokoko, thanks for the suggestions. I turned off auto update and it helps the cause. I didn't like that option in Creo either when it was introduced as a default. I turned it off there as well. I don't know why it didn't occur to me to do the same in Shark. I also enabled the deletion of undo. That seems to help a lot. Again thanks.

Cheers

Edited by user Sunday, January 21, 2018 5:58:55 PM(UTC)  | Reason: Not specified

one more...

Hey Murray, simply because I do not know yet what I am doing with Shark. I have been at it for about 14 days and that equates to Noob status in my book. I do not know how far I can go with the entire model. Meaning, I am not sure how many entities will push the limit for the code. Maybe it's unlimited, I just don't know yet. I was unsure if it was wise to build the large fore/aft surfaces as one and then split them. From what I have learned so far...I now know that is probably okay to do so. I was a bit apprehensive and unskilled. I have a friend that runs a code called IronCad and he is on a project where he has pushed it to it limit. Regens take too long and the code crashes frequently because too many calculations are overloading it. I think this was on my mind as I started my work. I think by the time I get a few more of these types of projects I will have a much better level of skill that will allow me to increase the finesse.

Edited by user Monday, January 22, 2018 7:33:17 PM(UTC)  | Reason: Not specified

Class A is an actual real endeavor. The following is a response I created on the Creo User Community Forum some time ago. It seems appropriate to copy it here.

**************************************************
True Class A modeling is the endeavor to capture design intent as closely and as accurately to the intent object as possible with numerical verification ability. Notice I said closely AND accurately. This is the point I made about the task being both subjective and explicit. The subjective aspect is what the you are "seeing" on the screen as you model. I.E the beautiful sweetened surfaces with nice highlights, zebra stripes, etc... These surfaces might look really awesome but might not be anywhere close to the actual intent (scan) surfaces created by the Designer. This is where Creo falls down with regard to Class A modeling. We have no tool, method or mechanism to easily ensure both the subjective AND the accuracy of our surfacing in Creo in the true sense of Class A modeling. The ability to precisely control G2 curvature nor anything regarding G3 acceleration and surface (local region and global) fullness (node manipulation) in all directions with smoothing/parsing is critical for Class A. Can we get close, lucky and or good with Creo? Yes. Can we create some stunning and cool stuff? Yes. However, Creo, and this includes Style and Freeform, is NOT, was NOT and never will be a Class A surfacing tool in the true sense of the term. We cannot fault the software for something it was never meant to be.

Creo "Class A" modeling is more a visual, subjective activity than it is a mathematically rigorous activity. True Class A is equal measures of both Art and Mathematics. We do not have good or easy control over curve degree, the ability to smooth/parse, add patches and then smooth them with G2 into the main surfaces, a highlight turntable, the list goes on... Typically, people that have a "Designers Eye" or training in Industrial Design will create surfaces with nuance and sweetness that most others will simply not consider but they will notice if its "missing". It's what I call the "Juice", it's that extra "something" that makes that particular thing really beautiful. When you get an opportunity go find a new vehicle sitting outside in the sunshine and start walking around it. Watch how the sunlight flows along the surfaces and does not wave or ripple. Pay attention to the creases in the body work and how they reflect the highlight. Look at the flow from one panel/component to another. All that work is done by a Class A modeler in a Class A tool like Catia, Alias, ICEM or NX. We do not have the ability to do that degree of work in Creo.

No bickering at all. I've done Class A on a number of projects for firms including Ford, Chrysler and Mack Trucks. Subjectivity only goes so far. There are actual numerical accuracy values I needed to achieve to even get into the Design Review meeting. If I was not good enough to get numerical approval, then I needed to revise it to get past the subjective approval. Are we arguing about silly things Murray? Let's not allow this to become some foolish pseudo argument. I am learning Shark, not surface development.

Hi N.T.Fla
Sorry, this is off topic, but I just had to throw this in.
This whole design project of yours is so taking me back 35 years ago to when I was a 1st year journeyman patternmaker.
The company I went to did jobbing work for various car companies.
A couple of projects I did were injection seat molds for the Holden Cameria(Aust) made out of fiberglass resin composite and aluminium frame, and some car front window testing jigs, The glass guys would make a window then throw it on the jig I had made to test them for shape size and curvature before shipping them out to the replacement windscreen guys.
Back then all we were given was a blueprint with a series of arcs and a datum line with 20 or so points to set your vernier height gauges to, then blend it in.
There was one guy there who was the most experienced, and he was working on the pattern for the plastic injection mold for motorised side mirror controllers, supposedly the first on an Australian manufactured car, I can't remember which one.
It was something that has stuck in my mind, he was casting with epoxy resin fine tapered (or drafted)2 to 3 mm fins, webs and bosses into the back of the curved mirror holder as well as maintaining a smooth and constant body thickness.
He was on it for at least 5 to 6 weeks. ( maybe he was sending the boss broke :-)..... )
Really it should have been done half by a toolmaker and half by a pattern maker but I think the boss was calling in favors as we needed the work.
I imagine its the type of job today they would do using a 5 axis cnc and a block of aluminum in a couple of days now, and dimensionly perfect to the drawings.
I remember there was a wall with about 50-100, 5, 6, 7, 800mm long french, bezier, body builder curves Hanging on hooks(they had a name but I cant recall what it was) made from Phenolic resin about 50mm wide. It wasn't a full set but quite a lot of the set.
They were some sort of dimensionaly accurate curves used by the model makers for shaping body panels.
I was instructed that I wasn't allowed to casually touch them as they were worth thousands, (I wonder what happened to them)

I lasted there about a year and then they brought in something called the John button car plan.
Basically the government decided that about 13 car manufactures in Aust was way too many and unsustainable, so the plan was to cut it back to 3 or 4, something like, Toyota, Holden, Ford and Mitsubishi.

Well that was the end of a whole lot of work for the boss infact the whole engineering industry took a huge downturn.
There was a recession also starting to hit.
Out of 7 pattermakers 4 of us went straight away, he kept the most experienced 3 and survived for a while.
The boss and his wife started selling imported toys from the front workshop to make ends meet. (absolutely no lie)
After I was put off I went about 4 weeks with no job, I then found a job process working in a ceramic tile factory for a while because any job was a good job. :-)

And thanks to the Button car plan and the government of the time, that's how you set a "blueprint" for how you go about dumbing down a nation.


But any way, yes I am looking at all the blended curves you are trying to implement thinking just smooth it out with a bit of sandpaper, or build it up with a bit of bog or make a template to compare one side with the other.

Meshes, lofted surfaces, skins, nurbs and such are rather such nebulous things don't you think?

However I suppose these days we need to know mathematically where every surface point of every panel is so it can be CNCed.

Loving your project.

Edited by user Tuesday, January 23, 2018 5:55:38 AM(UTC)  | Reason: Not specified

I have a mate that I do some jobbing for, a guy who manufactures truck mirrors in Sydney, CNC-from-a-billet for prototypes is it nowadays, but it wasn't always like that. He used to import from South America but that manufacturer wasn't too interested in his volume, he came back from Melbourne after convincing one of his clients to buy a style that didn't exist. They needed (a) local volume and (b) Australian Design Rules compliance, which mostly related to optical certification of prismatic reflectors for that thing. We mocked up a prototype over a weekend with some convex mirror glasses from another manufacturers' product, foam and body filler, rasps and fillet radius profiles to shape, the manufacturer committed, and we refined the design after the contract was signed. That was twelve or fifteen years ago, now he's manufacturing a full line of mirrors that go to a variety of OEM and aftermarket truck lines. Injection molds are ludicrously cheap these days compared to the early '80s Camira days, because of CAD and those CNC mills. The most interesting parts of developing them were designing the motorised gimbals with roll centres close to the glass surfaces and working out the movement envelope of glasses and mounting mechanisms within the bodies, which is surprisingly close to selfintersection problems in CAD, probably what your Camira guy was wrestling with, but he would've had to nut it out by building and experimenting.

The Button plan was designed to improve economies of scale, by encouraging (ie they got more money) manufacturers who built 40K of a single platform each year. If you're going to blame them for dumbing down the country by doing that, you've got to accept that it stems from reducing protectionism, which made imported cars cheaper by removing the tariffs that protected local jobs (in foreign-owned factories, not one Australian-owned), which made it harder and eventually impossible to build them here at competitive prices. Holden had 36K employees at its employment peak, and that many people built far fewer cars annually than the 5000 or so humans and the robots that they've had in recent years before they shut the doors.

It is sad, NeuTechFLA, but the world's changing, isn't it? BTW, do you know that Ford themselves are selling .stl models of some of their models through the TurboSquid content site for folks to 3D print their own miniatures at home? I bought one of them, a 2017 Focus RS, and as you'd expect, it's pretty nice and, also as you'd expect, it'd have to be the most accurate one out there. Sizing puts the model at about 1/28 scale, a new take on the promo models that dealers used to sell, and before that they even gave them away.

Hi Murray



Yes.
Cad and cheap CNC machines are decentralising what used to be the realm of large corporations.
Its one of the exciting aspects of life these days.
I think it will open up creativity again (which was stifled by the high cost of manufacturing during the 80's and 90's) for guys like you, me and your mate to capitalise on small niche markets.
Just think, items that would have taken teams of people and tons of money to get to market are now starting to become feasible for small 1,5, 10 man operations because of programs like viacad /shark and cheap cnc's



Yes. I understand.
((Its just that it was personal), (between me and Johny Button) :-)

But really I should stay on topic of N.T.FLA's interesting project.

A little Sunday progress. While I am "making progress", I still am on very shaky ground in reality with the software. Meaning, there is definitely ZERO finesse with what I am doing. I have no explicit G2 conditions anywhere and my spline transitions are...simply awful. And this is not really body design/surfacing at all. I am essentially creating Net Surfaces (in Creo they are called Boundary Surfaces) by playing a game of "connect the dots". Just here, I do not have the skill yet to use the G2 effectively. More time in the saddle will definitely help.

Observations:

A robust "untrim" would have come in really helpful in the front air dam insert area. Essentially, I am asking for feature History that can be accessed. But, I understand this adds "weight" to the code and makes things run slower.

Joining surfaces can, at times, result in one of the surfaces getting "trimmed" to some arbitrary shape. I seem to remember thin surfaces somewhat thinning to slivers at one boundary. Has this been seen by anyone before?