logo
NOTICE:  This is the new PunchCAD forum. You should have received an email with your new password around August 27, 2014. If you did not, or would like it reset, simply use the Lost Password feature, and enter Answer as the security answer.
Welcome Guest! To enable all features please Login or Register.

Notification

Icon
Error

2 Pages12>
Options
Go to last post Go to first unread
Art Smith  
#1 Posted : Sunday, May 6, 2018 2:31:09 PM(UTC)
Art Smith

Rank: Senior Member

Joined: 1/16/2015(UTC)
Posts: 351

Thanks: 49 times
Was thanked: 43 time(s) in 30 post(s)
process development for the ducts has progressed to the point joined sections of the duct surface can be "thickened"! using the "matching" tool set to position prior to joining the two sections was a must. given two "thickened" joined surfaces (ie: solids), I "added" the two together to form a single lumpy/crevassed solid.

filleting the resulting interface was the next step and I'm getting nowhere..... see attached image what I'd like to do is analogous to wiping an RTV fillet onto the part with your finger! the resulting interface from "adding" leaves a notch (ie: stress riser) and in my view is unsuitable for printing or casting. any thoughts on what I'm doing wrong or how to better proceed would be greatly appreciated!! SharkCAD Pro (1369) file attached. duct geometry is only representative for process development.

thanks!!

Art
[email protected]
File Attachment(s):
blending question.sfx (878kb) downloaded 5 time(s).
Art Smith attached the following image(s):
blending situation example.jpg (26kb) downloaded 4 time(s).

You cannot view/download attachments. Try to login or register.
murray  
#2 Posted : Sunday, May 6, 2018 10:51:42 PM(UTC)
murray

Rank: Senior Member

Joined: 9/24/2014(UTC)
Posts: 373
Australia

Thanks: 8 times
Was thanked: 104 time(s) in 85 post(s)
Are these mirrored? One of the branches has imperfections around its rounded-rectangular end, so I changed the part into surfaces and thickened the "good" inner surfaces to .12, mirrored it, added them and filleted.
File Attachment(s):
blending question mod.sfx (1,999kb) downloaded 3 time(s).
murray attached the following image(s):
Duct theathon.PNG (186kb) downloaded 3 time(s).

You cannot view/download attachments. Try to login or register.
thanks 1 user thanked murray for this useful post.
Art Smith on 5/6/2018(UTC)
Art Smith  
#3 Posted : Sunday, May 6, 2018 11:34:12 PM(UTC)
Art Smith

Rank: Senior Member

Joined: 1/16/2015(UTC)
Posts: 351

Thanks: 49 times
Was thanked: 43 time(s) in 30 post(s)
murray-

as best I can remember (ie: configurations come and go when things aren't working as expected), they're not "mirrored", but very close relatives! imperfections ??? where exactly AND how were they detected/found?? my number one problem with SharkCAD Pro is the useless "error messages" and complete lack of "clues" when there's a problem that needs fixing........... your fillet is exactly what I'm attempting to make happen!! thanks

Art
[email protected]

murray  
#4 Posted : Monday, May 7, 2018 1:03:11 AM(UTC)
murray

Rank: Senior Member

Joined: 9/24/2014(UTC)
Posts: 373
Australia

Thanks: 8 times
Was thanked: 104 time(s) in 85 post(s)
It's interpretation, not an error message. One of the rounded-rectangular ends wouldn't take a cover surface because it sees the curves (edges of the surfaces) as open. It's still not altogether workable, because when I try to close it off and stitch it into a solid for re-shelling to accommodate the blend, stitching fails. I'll mess with it later on and see what happens.
Art Smith  
#5 Posted : Monday, May 7, 2018 8:37:08 AM(UTC)
Art Smith

Rank: Senior Member

Joined: 1/16/2015(UTC)
Posts: 351

Thanks: 49 times
Was thanked: 43 time(s) in 30 post(s)
following-up on murray's observation of "imperfections", I started poking at the geometry and don't understand what's going on. four surfaces are in play; 2 "Guide Cover Surfaces" and 2 "Loft Surfaces". both "Guide Cover Surfaces" appear to both "Match (position)" & "Join" AND both "Loft Surfaces" appear to both "Match (position)" & "Join". unexpected behavior results from attempting to "Match" either the matched & joined guide cover surfaces to the matched & joined lofted surfaces OR the matched & joined lofted surfaces to matched & joined guide cover surfaces !! it's my expectation that attempting to match two surfaces shouldn't result in loosing either of the surfaces..........
>>> what's going on with "Matching" ??
>>> is it attempting to tell me something ??
>>> and if so, what ??
>>> and is this related to my inability to fillet the next level assembly (see previous post above) ??

collage and file attached below. thanks!

Art
[email protected]
File Attachment(s):
blending question 2.sfx (887kb) downloaded 3 time(s).
Art Smith attached the following image(s):
blending situation example 2.jpg (46kb) downloaded 6 time(s).

You cannot view/download attachments. Try to login or register.
rockyroad_us  
#6 Posted : Monday, May 7, 2018 9:07:22 AM(UTC)
rockyroad_us

Rank: Senior Member

Joined: 9/12/2009(UTC)
Posts: 600
Man
United States
Location: In my place

Was thanked: 52 time(s) in 46 post(s)
Looks to me like you have an upper area that overlaps and an opposite area that has a gap. I would stitch the yellow and the green to make solids and use the match face of the yellow to the green or vice versa.

But maybe the cross-sections are not the same.
FX 7-FXPro14-1650
MacPro 11.7.4
2x2.8GHz Quad Xeon 32GB
NVIDIA 4 K5000 4GB
Work
murray  
#7 Posted : Monday, May 7, 2018 5:04:43 PM(UTC)
murray

Rank: Senior Member

Joined: 9/24/2014(UTC)
Posts: 373
Australia

Thanks: 8 times
Was thanked: 104 time(s) in 85 post(s)
I'd try to avoid frustration deriving profiles from the work you've already done and model it solid. How is it intended to be molded, Art? It suggests blow molding to me, but the mold pieces and actions don't look easy with the branches converging and the floor between them. You've said that it's representative, not definitive, but it looks like a V-twin intake filter duct to me. Doesn't strike me as a bike engine aesthetic, but maybe a compressor or industrial engine?

Edited by user Monday, May 7, 2018 6:20:11 PM(UTC)  | Reason: Not specified

murray  
#8 Posted : Monday, May 7, 2018 6:33:25 PM(UTC)
murray

Rank: Senior Member

Joined: 9/24/2014(UTC)
Posts: 373
Australia

Thanks: 8 times
Was thanked: 104 time(s) in 85 post(s)
An observation, but I opened your original posted file and it will fillet for me. It's using inches and the blend is .3937 ie 10mm. Lots of material in the duct walls where it is filleted, though, lots of sink if molded like that.
murray attached the following image(s):
original.PNG (78kb) downloaded 1 time(s).

You cannot view/download attachments. Try to login or register.
Art Smith  
#9 Posted : Monday, May 7, 2018 7:53:17 PM(UTC)
Art Smith

Rank: Senior Member

Joined: 1/16/2015(UTC)
Posts: 351

Thanks: 49 times
Was thanked: 43 time(s) in 30 post(s)
murray-

thanks! just tried the posted file and blending now works for me...........?? go figure as I haven't done a thing to the file!

"How is it intended to be molded, Art? It suggests blow molding to me, but the mold pieces and actions don't look easy with the branches converging and the floor between them. You've said that it's representative, not definitive, but it looks like a V-twin intake filter duct to me. Doesn't strike me as a bike engine aesthetic, but maybe a compressor or industrial engine?"

a friend asked me to take a look at what might be possible for a Weber 55mm DCO (not the DCOE with chokes) crossram manifold for an extended deck SBC. seeing the manifold design as both an extension of the work I've been doing on cylinder heads AND an opportunity to use the surface creation/modification tools that now dominate discussion on the forum, I agreed to take a look. implicit in taking a look would be my aerospace top-down requirements driven approach focused on engineering/physics and not appearance. 300 ft/sec intake flow rates are effectively equivalent to 200 mph where everyone understands aerodynamics is important!! attached is an image of an early version of the only option considered with exactly the same duct/port for ALL eight cylinders; the combined geometry constraints from the Weber and the two cylinder heads is a nightmare (but doable) !!! the resulting aspect ratio for the duct/port was 5 2/3 : 1 (ie: 4.25/0.75), well beyond everything else I've specified to date. in the end he decided to go with different length ducts/ports to get shorter runners with a more typical aspect ratio......... lefts AND rights doubles the work AND requires all the little things have to be spot-on so as not to aggravate the difference in runner lengths. at the end of the day it won't be hard to do better than the 40mm Weber
manifold he has given the obvious difference in its runner lengths AND inability to handle the 55mm DCO's!!

carbon fiber reinforced "printed" plastic is the stated objected for reduced weight AND heat transfer. one of my unspoken assignments is to preserve both a thin wall aluminum casting option and at least one forming option.............................


Art
[email protected]
Art Smith attached the following image(s):
completing the illusion.jpg (34kb) downloaded 4 time(s).

You cannot view/download attachments. Try to login or register.
murray  
#10 Posted : Monday, May 7, 2018 9:23:04 PM(UTC)
murray

Rank: Senior Member

Joined: 9/24/2014(UTC)
Posts: 373
Australia

Thanks: 8 times
Was thanked: 104 time(s) in 85 post(s)
Wow, pretty old school. I worked for a speed/karting shop for a couple of years forty years ago while I was playing with bikes and karts, the place had DCOEs, other sidedraft and downdraft Webers and Dellortos everywhere, Solexes and Mikunis for Brit and Japanese sporties, more Mikunis and other flat-slide carbs for two-stroke bikes and karts. Plus the manifolds most-wanted, in New Zealand back then they were old BMC Mini engines, Ford Kent engines like Formula Ford 1600, occasional Fiat and Lotus twin-cams, Holden (Australian GM) and Australian Ford sixes that were the most popular of the day. Quite a few of those had siamese inlet ports, but they used to make a big deal about straight-run paths and use that to sell manifolds that used two DCOEs or Dellortos, but with one barrel of each blanked off and the other barrel leading straight into the horribly-compromised head casting, because people would go all quiet and dreamy-eyed seeing these manifolds and open their wallets to buy the two carbs.... BTW, has your friend heard of "fuel injection"? kidding... Considering the proportion of time that engines spend WOT at sustained speeds, isn't the "equal length, constant area cross-section" spec sort of theoretically fascinating and in practice almost entirely irrelevant?

Edited by user Monday, May 7, 2018 9:32:40 PM(UTC)  | Reason: Not specified

Art Smith  
#11 Posted : Monday, May 7, 2018 9:43:51 PM(UTC)
Art Smith

Rank: Senior Member

Joined: 1/16/2015(UTC)
Posts: 351

Thanks: 49 times
Was thanked: 43 time(s) in 30 post(s)
murray-

the oldies but goodies (ie: Webers) keep popping up......... it's my sense avoiding electronics and the need for a computer is motivating the "take no prisoners" Webers approach. 55mm is 2.165" which means done properly there's a full throat path to the cylinder; remember SBC intake valves start at 2.1"

as for the impact of runner length variation, it's the #1 intake problem with my FFord Kent engines!! the inner pair and the outer pair require different jetting AND creative intake manifold porting to get the EGT values approximately the same in the power band!! and at least in Kent's, minimal EGT variation is the key to optimal ignition timing

Art
[email protected]
murray  
#12 Posted : Monday, May 7, 2018 9:59:32 PM(UTC)
murray

Rank: Senior Member

Joined: 9/24/2014(UTC)
Posts: 373
Australia

Thanks: 8 times
Was thanked: 104 time(s) in 85 post(s)
Have you heard the one about how to get 200 bhp from an FF Kent, Art? Back in the day, there was a legendary engine that won a few festivals in the UK, and it was reputed to have been built from a block that had been buried in the ground, "seasoned" as the builder put it, for a decade or more, and some other builders' methods were equally mythic. I forget who the joke was attributed to originally, but the trick is to get a good engine to begin with, then take it to a rival builder, who'll solicit your business by dismissing the original builder's efforts with something like "Him? I can get you 5% more power if you give it to me..." So he does your next rebuild, then you take it to the next feller, who'll tell you approximately the same thing. Following that pattern, eventually, you'll get your 200 brake...
Art Smith  
#13 Posted : Wednesday, May 9, 2018 8:48:54 PM(UTC)
Art Smith

Rank: Senior Member

Joined: 1/16/2015(UTC)
Posts: 351

Thanks: 49 times
Was thanked: 43 time(s) in 30 post(s)
does anyone have any experience/thoughts regarding elimination of ripples resulting from the use of the "Guide Loft Surface" tool?? after building station sections on the order of every 1/4" of runner length, the "Guide Loft Surface" tool is producing surfaces that are almost usable. see attached collage. given the small difference in inlet to outlet area AND short length, I decided to go with the linear area approach simplifying things immensely. still a good deal of spreadsheet work but nothing compared to implementing a exponential horn with a fixed width.........

clearly I'm willing to consider ALL the usual suspect:
more stations
more guide curves
more compute intensive option/tool; assume anything less than 12 hours is on the table!
exporting to another (at this point unknown) application and then importing the desired surface

short side radius ripples are a non-starter for both a near-net printing and investment casting implementation................ thanks

Art
[email protected]
Art Smith attached the following image(s):
lofting ripples question.jpg (76kb) downloaded 12 time(s).

You cannot view/download attachments. Try to login or register.
UGMENTALCASE  
#14 Posted : Thursday, May 10, 2018 2:58:55 AM(UTC)
UGMENTALCASE

Rank: Senior Member

Joined: 3/21/2017(UTC)
Posts: 945
Man
United Kingdom

Thanks: 46 times
Was thanked: 298 time(s) in 203 post(s)
Do you have the file?
Art Smith  
#15 Posted : Thursday, May 10, 2018 11:00:50 AM(UTC)
Art Smith

Rank: Senior Member

Joined: 1/16/2015(UTC)
Posts: 351

Thanks: 49 times
Was thanked: 43 time(s) in 30 post(s)
file without lofted surface 391 KB
file with lofted surface 2.11 MB

Art
[email protected]
File Attachment(s):
lofting ripple question file.sfx (2,168kb) downloaded 6 time(s).

You cannot view/download attachments. Try to login or register.
rockyroad_us  
#16 Posted : Thursday, May 10, 2018 11:23:36 AM(UTC)
rockyroad_us

Rank: Senior Member

Joined: 9/12/2009(UTC)
Posts: 600
Man
United States
Location: In my place

Was thanked: 52 time(s) in 46 post(s)
Looks to me like your lines are not intersecting properly. For instance at the area of the circle, the lines should end up exactly at the vertexes. another thing is that the guide curves should have a point where they intersect the rectangular profiles. That has been my experience using Shark to get a smooth surface as much as possible.

FX 7-FXPro14-1650
MacPro 11.7.4
2x2.8GHz Quad Xeon 32GB
NVIDIA 4 K5000 4GB
Work
Art Smith  
#17 Posted : Thursday, May 10, 2018 12:17:44 PM(UTC)
Art Smith

Rank: Senior Member

Joined: 1/16/2015(UTC)
Posts: 351

Thanks: 49 times
Was thanked: 43 time(s) in 30 post(s)
rockyroad_us-

thanks for your input. as time permits, a couple of screen shots would help me understand your observations.

as for the four guide curves, the spline curves were created with points from each of the station rectangles; ie: the reason for creating the local x-axis and local y-axis. additional guide curves could be created expediently using the ends of the corner arcs that make up the station IF there's potential merit!

Art
[email protected]
UGMENTALCASE  
#18 Posted : Thursday, May 10, 2018 2:16:38 PM(UTC)
UGMENTALCASE

Rank: Senior Member

Joined: 3/21/2017(UTC)
Posts: 945
Man
United Kingdom

Thanks: 46 times
Was thanked: 298 time(s) in 203 post(s)
Take a look at the attached. Had a little play with it and got half of the profile looking good. Then repeat for the other side and join together?

Cheers
File Attachment(s):
lofting ripple question file_trimmed curves.slt (2,127kb) downloaded 4 time(s).

You cannot view/download attachments. Try to login or register.
thanks 1 user thanked UGMENTALCASE for this useful post.
Art Smith on 5/10/2018(UTC)
Art Smith  
#19 Posted : Thursday, May 10, 2018 2:43:10 PM(UTC)
Art Smith

Rank: Senior Member

Joined: 1/16/2015(UTC)
Posts: 351

Thanks: 49 times
Was thanked: 43 time(s) in 30 post(s)
Originally Posted by: UGMENTALCASE Go to Quoted Post
Take a look at the attached. Had a little play with it and got half of the profile looking good. Then repeat for the other side and join together?

Cheers



UGMENTALCASE-

thanks, looks outstanding!! two question: 1.) "how" ?? and 2.) is your approach to doing the surface in pieces likely to be compatible with converting the resulting collection of surfaces to a solid AND then blending two or more of the solids? I started down the lofting road because it has the potential to yield a single surface compatible with Shark's other tools without the error messages encountered in the past trying to use matching, joining, and blending...................

Art
[email protected]
UGMENTALCASE  
#20 Posted : Thursday, May 10, 2018 3:32:28 PM(UTC)
UGMENTALCASE

Rank: Senior Member

Joined: 3/21/2017(UTC)
Posts: 945
Man
United Kingdom

Thanks: 46 times
Was thanked: 298 time(s) in 203 post(s)
Ha ha, how?! Well It doesn't like the squished bottom and expanded top when you use some of the surface options.
So I used the option where you pick M and N curves. But I found the results changed when you selected either the guide curves first or the other curves first. So in this case I selected the guide curve type, and then the 7 or 8 profile curves.

I'd say you can adjust them to solid at the end yes. Can't see any reason why not. Repeat my steps to build the other half of that section. Then another surface to go from the circle shape to the rectangular type shape, join/add them all togetber, then thicken up maybe?

I'll build the other half in the morning, then run in he other surface and test it out. How thick do you need it?
Users browsing this topic
Guest
2 Pages12>
Forum Jump  
You cannot post new topics in this forum.
You cannot reply to topics in this forum.
You cannot delete your posts in this forum.
You cannot edit your posts in this forum.
You cannot create polls in this forum.
You cannot vote in polls in this forum.