Keep in mind this is an open debate, .. play nice guys
as the debate between you will add great depth to this thread.
This is getting really good !!
Chuck does make a good point about the economics of the problem, ..
and the customer's perception.
OK, .. Jon you make a few very good and serious points.
Let say for instance that I had a scan or file of a 15º chevy head, ..
for an engine that needed 2.6" of minimum cross sectional area, ..
and a customer needs a head with a CSA of 3", .. the ability to
"open" the port in the file and cut it VS having rescan a port that is
the proper size would be a benefit.
Having a shape that works and only needs more cross section for a different
application, .. happens all the time.
Now I know nothing about Geomagic outside of the videos & stuff I got
from their web site. It looks like software I could learn very fast, .. and I can easly
see it does what I want, .. can this software "enlarge" a scanned port?
Can't Geomagic handle those "lumps" as well???
I'm learning this stuff slow, Surfcam isn't very intuitive, .. but I can see how
trying to take a modeled port in Surfcam, and increase the cross sectional
area would be a major pain, .. .....
I think this may lead back to my question, .. .. should there be two work flows
depending on the project??
I'm going to try and get some heads scanned next week, .. unless some one has
some point clouds or scans they'd like to offer for our debate.
Curtis
CnC head software ?
Moderator: Team
-
SchmidtMotorWorks
- Vendor
- Posts: 11003
- Joined: Mon Apr 11, 2005 2:30 am
- Location: CA
Chuck does make a good point about the economics of the problem, ..
and the customer's perception.
Not really,
As for the economies, I have been using CAD CAM since 1979 and have watched one industry after another company by company either take advantage of CAD CAM to the fullest or go out of business. Just a few years ago it was cosidered ridiculous to shape surfboards with CNC equipment now non CNC boards are considered second rate. The list of transformed industries seems endless yet I can't think of one that has reversed the trend.
A lot of motorsports people always look for the low cost solution rather than the high ROI solution at their own expense.
As for marketing the advantgaes of using a technology whos result cannot be exactly seen is not hard to find, plenty of heads are now marketed on the characteristic of beign CNC ported.
To follow Chucks argument, it wouldn't make sense to advertise that the heads are CNC ported, that is unless you make the case for why they are better. In the current context CNC are considered better by many because they are consistant with a supposedly good design.
There is ample evidense that objects that are properly engineered with CAD will be superior over time to those that are not, to argue against this is foolish. The work for anyone in marketing to do is to educate thier customers why what they are doing produces a superior product and ride the advantage as long as you can hold on to your technical lead.
Since port modeling is really difficult you might have some time because many will try and fail. I watched this happen in the Mold Industry in the early 90s when Pro E first came out. Every shop around bought it with big dreams of doing molds in half the time or less but only a small percent had the stamina to work through all the problems and develope a complete connected workable process. Those that did are the only ones left standing to fight against Chinese competition, who by the way adopt CAD CAM with much less resistance than Americans do.
-
SchmidtMotorWorks
- Vendor
- Posts: 11003
- Joined: Mon Apr 11, 2005 2:30 am
- Location: CA
Nope, Geomagic uses lots of small surface patches and quilts them together and all the surfaces must have four sides.Can't Geomagic handle those "lumps" as well???
To give an analogy, imagine fabricating someting like the shape of a port with a pile of postage stamp sized pieces of sheet metal. The result is very similar.
About four years ago I was tasked with evaluating variuos Reverse Engineering software. I spent about 6 months using about 8 different softwares to model a hood and fender from scanned points or facets whatever the software required. None of them produced direcltly usable surfaces even with the top AEs from each company showing how to use it. At best I had a good reference to use for a parametric model.
Another project I worked on a couple of years ago was testing a mold cavity splitting software on about 300 different plastic parts from all over the world everyting from dashboards to hairdryers you name it modeled in every software I ever heard of and more. One thing that I remember about that test is that we were looking for examples of parts that were produced from Reverse Engineering software to use in the testing. Not one of the mold shops we had as customers had made a mold for a part that was designed for production with a Reverse Eningineering software.
If it could really do what they claim, there would be more products designed with it.
That said, it would probably work fine as a volume for FEA as most people would mesh an engine block at sizes above 0.1" so innacuracies wouldn't matter.
Schmidt, thanks for your post, I have learned a considerable amout in a short time. In one of the post someone questioned about the difficulty of raising the roof on a model. What about taking an existing port design and scaling it up for a different bore size. Example, taking a W9RP port shape and enlarging it to fit a BB Mopar. Ford used the SB Yates head as a template for the C460. The ports were enlarged to fit the needs of the bigger engine. Can this be done? 
-
SchmidtMotorWorks
- Vendor
- Posts: 11003
- Joined: Mon Apr 11, 2005 2:30 am
- Location: CA
Scaling non parametric data to get a new design is a tricky issue. If you really mean "scaling", in CAD software that term describes an operation that really does exactly that, you select the part you want to scale and enter a scale value like 1.25 to increase the size by 25%. Sounds easy right? Now you go stick the port in a head and find a list of problems like the following:
1. The valve angle may be different
2. The distance from the valve origin to the intake face is different.
3. The intake face angle is different
4. The level of the intake opening to the seat are different
5. The position of the pushrod pinch point is different
6. The offset from the cylinder centerline to the valve centerline is different
7. The distance from the port opening to the adjacent port is different.
OK so now you have problem that would give asprin a headache.
So you can try a trick, some software can scale different ammounts in each axis, this might solve some of the problems but it is just as likely to cause new problems. When you scale in non uniform ways you change cylindrical shapes to elliptical shapes, everything gets sort of distorted. Now your bowl doesn't match the seat any more.
So this is the beauty of a parametric model. All of those problems in a parametric model have corresponding parameters that can be edited from a list.
Just to give youan idea I have posted part of a list from an existing head i am working on.
This list is typical for a cylider layaout set, there are other sets for the head layout (an assembly of mutiple cyliders), and the engine layout, all these are kind of meshed together with related dimensions.
The expressions with name = simple numeric value are the controlling dimensions.
The expressions with name= file name:expression name are expressions that are related form diffeent layout levels of the engine.
The expressions that a p= expression are the parameters of geometric features that are controlled by those expressions.
bore_dia=cyl_head_layout_assy::bore_dia
bore_spacing=engine_layout_assy::bore_spacing
cam_base_rad=.625
cam_level=bore_dia*1.05
cam_roller_rad=.5
cam_valve_offset=2.2
deck_ex=-engine_layout_assy::deck_ex
deck_in=engine_layout_assy::deck_in
ex_face_angle=10
ex_face_z_start_level=1.4
ex_port_turn_dist=2.5
ex_port_y=-5
ex_rocker_pivot_drop=rocker_pivot_drop_factor*ex_valve_lift
ex_seat_ht=.3
ex_seat_id=ex_seat_od-.3
ex_seat_od=ex_valve_dia+.1
ex_spring_level=3.25
ex_valve_angle_y=15
ex_valve_dia=bore_dia*.333
ex_valve_lift=1
ex_valve_long=valves_long
ex_valve_origin_x=bore_dia*.21
ex_valve_origin_y=-bore_dia*.23
ex_valve_origin_z=bore_dia*0.076
in_face_angle=10
in_face_y=5
in_face_z_start_level=1.3
in_port_turn_dist=2.5
in_port_y=5
in_rocket_pivot_drop=rocker_pivot_drop_factor*in_valve_lift
in_seat_ht=.3
in_seat_id=in_seat_od-.3
in_seat_od=in_valve_dia+.1
in_spring_level=3.25
in_valve_angle_y=15
in_valve_dia=bore_dia*.4
in_valve_lift=1
in_valve_long=valves_long
in_valve_origin_x=bore_dia*.22
in_valve_origin_y=bore_dia*.17
in_valve_origin_z=bore_dia*.065
lash_cap_thick=0.04
lifter_bore_top_drop=.75
p0=bore_dia
p1=in_valve_origin_x
p2=in_valve_origin_y
p3=in_valve_origin_z
p4=in_valve_angle_y
p6=90
p8=ex_valve_origin_x
p9=ex_valve_origin_y
p10=ex_valve_origin_z
p11=-ex_valve_angle_y
p12=90
p13=in_seat_ht
p14=ex_seat_ht
p17=deck_in
p18=deck_ex
p25=ex_spring_level
p26=in_spring_level
p27=(in_valve_angle_y-ex_valve_angle_y)/2
p29=bore_spacing/2
p30=-bore_spacing/2
p31=2.25
p57=1.4
p63=2.3
p64=1.9
p65=1.75
p66=1.75
p83=spark_origin_y
p84=spark_origin_z
p85=-spark_angle_y
p86=90
p94=cam_level
p95=cam_level+rocker_pivot_offset_z
p98=-ex_valve_origin_x
p99=-in_valve_origin_x
p100=2.6
p101=2.5
p102=-0.60
p113=in_face_y
p114=ex_port_y
p117=ex_face_angle
p118=-in_face_angle
p120=90
p121=90
port_level=bore_spacing*.4
rocker_pivot_drop_factor=.5
rocker_pivot_offset=3.2
rocker_pivot_offset_z=bore_dia*.375
rocker_pivot_to_tip_roller_dist=rocker_pivot_offset+.02
rocker_ratio=1.8
rocker_roller_offset=rocker_pivot_offset-(rocker_pivot_offset/rocker_ratio)
rocker_valve_roller_rad=.375
spark_angle_y=-15
spark_origin_y=-.175
spark_origin_z=.5
valves_long=6.0
This is the valve layou it controls
1. The valve angle may be different
2. The distance from the valve origin to the intake face is different.
3. The intake face angle is different
4. The level of the intake opening to the seat are different
5. The position of the pushrod pinch point is different
6. The offset from the cylinder centerline to the valve centerline is different
7. The distance from the port opening to the adjacent port is different.
OK so now you have problem that would give asprin a headache.
So you can try a trick, some software can scale different ammounts in each axis, this might solve some of the problems but it is just as likely to cause new problems. When you scale in non uniform ways you change cylindrical shapes to elliptical shapes, everything gets sort of distorted. Now your bowl doesn't match the seat any more.
So this is the beauty of a parametric model. All of those problems in a parametric model have corresponding parameters that can be edited from a list.
Just to give youan idea I have posted part of a list from an existing head i am working on.
This list is typical for a cylider layaout set, there are other sets for the head layout (an assembly of mutiple cyliders), and the engine layout, all these are kind of meshed together with related dimensions.
The expressions with name = simple numeric value are the controlling dimensions.
The expressions with name= file name:expression name are expressions that are related form diffeent layout levels of the engine.
The expressions that a p= expression are the parameters of geometric features that are controlled by those expressions.
bore_dia=cyl_head_layout_assy::bore_dia
bore_spacing=engine_layout_assy::bore_spacing
cam_base_rad=.625
cam_level=bore_dia*1.05
cam_roller_rad=.5
cam_valve_offset=2.2
deck_ex=-engine_layout_assy::deck_ex
deck_in=engine_layout_assy::deck_in
ex_face_angle=10
ex_face_z_start_level=1.4
ex_port_turn_dist=2.5
ex_port_y=-5
ex_rocker_pivot_drop=rocker_pivot_drop_factor*ex_valve_lift
ex_seat_ht=.3
ex_seat_id=ex_seat_od-.3
ex_seat_od=ex_valve_dia+.1
ex_spring_level=3.25
ex_valve_angle_y=15
ex_valve_dia=bore_dia*.333
ex_valve_lift=1
ex_valve_long=valves_long
ex_valve_origin_x=bore_dia*.21
ex_valve_origin_y=-bore_dia*.23
ex_valve_origin_z=bore_dia*0.076
in_face_angle=10
in_face_y=5
in_face_z_start_level=1.3
in_port_turn_dist=2.5
in_port_y=5
in_rocket_pivot_drop=rocker_pivot_drop_factor*in_valve_lift
in_seat_ht=.3
in_seat_id=in_seat_od-.3
in_seat_od=in_valve_dia+.1
in_spring_level=3.25
in_valve_angle_y=15
in_valve_dia=bore_dia*.4
in_valve_lift=1
in_valve_long=valves_long
in_valve_origin_x=bore_dia*.22
in_valve_origin_y=bore_dia*.17
in_valve_origin_z=bore_dia*.065
lash_cap_thick=0.04
lifter_bore_top_drop=.75
p0=bore_dia
p1=in_valve_origin_x
p2=in_valve_origin_y
p3=in_valve_origin_z
p4=in_valve_angle_y
p6=90
p8=ex_valve_origin_x
p9=ex_valve_origin_y
p10=ex_valve_origin_z
p11=-ex_valve_angle_y
p12=90
p13=in_seat_ht
p14=ex_seat_ht
p17=deck_in
p18=deck_ex
p25=ex_spring_level
p26=in_spring_level
p27=(in_valve_angle_y-ex_valve_angle_y)/2
p29=bore_spacing/2
p30=-bore_spacing/2
p31=2.25
p57=1.4
p63=2.3
p64=1.9
p65=1.75
p66=1.75
p83=spark_origin_y
p84=spark_origin_z
p85=-spark_angle_y
p86=90
p94=cam_level
p95=cam_level+rocker_pivot_offset_z
p98=-ex_valve_origin_x
p99=-in_valve_origin_x
p100=2.6
p101=2.5
p102=-0.60
p113=in_face_y
p114=ex_port_y
p117=ex_face_angle
p118=-in_face_angle
p120=90
p121=90
port_level=bore_spacing*.4
rocker_pivot_drop_factor=.5
rocker_pivot_offset=3.2
rocker_pivot_offset_z=bore_dia*.375
rocker_pivot_to_tip_roller_dist=rocker_pivot_offset+.02
rocker_ratio=1.8
rocker_roller_offset=rocker_pivot_offset-(rocker_pivot_offset/rocker_ratio)
rocker_valve_roller_rad=.375
spark_angle_y=-15
spark_origin_y=-.175
spark_origin_z=.5
valves_long=6.0
This is the valve layou it controls
I think I need my Buck Rogers secret decoder ring for that. 
Just kidding.Thanks schmidt for the answer. ShaneThis is getting really, really interesting, .. 
Jon,
So in basic terms, .. if I was to use a 3 axis or 4 axis machine, ..
and wanted to reverse engineer my ports, .. touch probe scan, ..
what software solutions would I use?
What would be the easiest to learn & use for some one not
so "involved" with this type of software, .. and what would the
"best" solution be that didn't take the user's skill into account?
I wonder, .. I know many are using Surfcam, .. scan with a Renishaw
Cyclone, .. what is the work flow there?
It seems that the "limited" solution with Surfcam is common where if
a change to a port design is required a new scan is also required.
Just how have others solved this problem? Chapman's CnC work
is beautiful, Weldtech does a nice job, .. Fultz's, .. etc, .. what have
they done?
Something else that really hasn't been touched on is the expense of the different
software packages. $30K plus a yearly maintenance fee is common for this stuff, ..
it's very expensive.
If you have a work flow that uses Geomagic, into Pro-e as an example, ..
you've got over $60K in software plus $3-4K in yearly maintenance, ..
VERY cost prohibitive for a small shop.
So again, .. it seems like there's a less expensive way to do this but has
big limitations, .. and the good way to do this, .. but expensive, .. true?
As far as Chuck's evaluation of the CnC cylinder head business I have to agree
with Chuck, being I deal with the customer's every day. CnC has been "advertised"
as being better because it's more consistent and can duplicate a proven design.
Most customers ask if I CnC heads, .. they don't know why they ask other then
they read in magazines it's the best thing.
The bottom line is does it make power & win races, and most customers
don't care how I do that, .. the technology being sold is the cylinder head
development not the tools being used to produce it.
The truth is in today's current cylinder head development a CnC machine
is nothing more then a big copy machine. If a bad port design is scanned
and duplicated on CnC, there's not magic HP button on the CnC machine,
all it'll do is duplicate the same bad port design.
So, the hand ported head, .. that has been tested on the dyno & race track, ..
has to come first, .. and as far as I'm concerned always will.
I'd only scan & CnC a port I know works, ..
Now my intended initial use of CnC is much like Shawn has brought up.
I want a machine to "rough" ports that are then hand finished to the specs
of the project. Throat size, .. bowl size, .. short turn cross section all change
as per the app. It's much more productive to take an hour to hand work
a valve bowl for the intended use then to spend several hours in software
making changes.
Thanks for all the input, .. again guys, .. great thread!
Curtis
Jon,
So in basic terms, .. if I was to use a 3 axis or 4 axis machine, ..
and wanted to reverse engineer my ports, .. touch probe scan, ..
what software solutions would I use?
What would be the easiest to learn & use for some one not
so "involved" with this type of software, .. and what would the
"best" solution be that didn't take the user's skill into account?
I wonder, .. I know many are using Surfcam, .. scan with a Renishaw
Cyclone, .. what is the work flow there?
It seems that the "limited" solution with Surfcam is common where if
a change to a port design is required a new scan is also required.
Just how have others solved this problem? Chapman's CnC work
is beautiful, Weldtech does a nice job, .. Fultz's, .. etc, .. what have
they done?
Something else that really hasn't been touched on is the expense of the different
software packages. $30K plus a yearly maintenance fee is common for this stuff, ..
it's very expensive.
If you have a work flow that uses Geomagic, into Pro-e as an example, ..
you've got over $60K in software plus $3-4K in yearly maintenance, ..
VERY cost prohibitive for a small shop.
So again, .. it seems like there's a less expensive way to do this but has
big limitations, .. and the good way to do this, .. but expensive, .. true?
As far as Chuck's evaluation of the CnC cylinder head business I have to agree
with Chuck, being I deal with the customer's every day. CnC has been "advertised"
as being better because it's more consistent and can duplicate a proven design.
Most customers ask if I CnC heads, .. they don't know why they ask other then
they read in magazines it's the best thing.
The bottom line is does it make power & win races, and most customers
don't care how I do that, .. the technology being sold is the cylinder head
development not the tools being used to produce it.
The truth is in today's current cylinder head development a CnC machine
is nothing more then a big copy machine. If a bad port design is scanned
and duplicated on CnC, there's not magic HP button on the CnC machine,
all it'll do is duplicate the same bad port design.
So, the hand ported head, .. that has been tested on the dyno & race track, ..
has to come first, .. and as far as I'm concerned always will.
I'd only scan & CnC a port I know works, ..
Now my intended initial use of CnC is much like Shawn has brought up.
I want a machine to "rough" ports that are then hand finished to the specs
of the project. Throat size, .. bowl size, .. short turn cross section all change
as per the app. It's much more productive to take an hour to hand work
a valve bowl for the intended use then to spend several hours in software
making changes.
Thanks for all the input, .. again guys, .. great thread!
Curtis
Race Flow Development
Simultaneous 5-axis CNC Porting
http://www.raceflowdevelopment.com
Simultaneous 5-axis CNC Porting
http://www.raceflowdevelopment.com
-
SchmidtMotorWorks
- Vendor
- Posts: 11003
- Joined: Mon Apr 11, 2005 2:30 am
- Location: CA
I'm just heading out the door to go to Ohio to look for a new home and shop, until tuesday so this will it for now.
Lots of companies are modeling and producing ports that have never been touched with a die grinder.
At GM, they make these plastic 2 piece boxes they make hundreds in advance. They create CAD models and cut the shape into the port boxes (unattended) which they test on a flow bench. One person working alone could easily modify a port design to four port variants and cut port boxes in one day. You can multi task by cutting the ports and flow testing at the same time.
Just get a good probe and digitizing software, or maybe a low end coordinate measuring device. Learn to model parametric ports, cut the shape into some port boxes until you get something you like. Cut that into some heads. Make some adjustments if it doesn't fit the casting, grind what you can't reach.
I think you may want to consider removing the hand ported port from the first stages of port design, over time you will find it to be easier and less time with a die grinder in your hands. After reading about the suffering the die grinder causes, I would think you all would be walking over hot coals to escape the die grinder time.
After a few ports have been done the way I am telling you, you will wonder why you ever would start a protoype with a die grinder. When surface modeling becomes as familiar to you as grinding your perspective on this will change.
When I get back, I will try to make a series of images and maybe some video of port modeling in CAD. I think the unkown aspect of this is making this difficult to accept. Seeing an exact step by step example should clear that up.
Lots of companies are modeling and producing ports that have never been touched with a die grinder.
At GM, they make these plastic 2 piece boxes they make hundreds in advance. They create CAD models and cut the shape into the port boxes (unattended) which they test on a flow bench. One person working alone could easily modify a port design to four port variants and cut port boxes in one day. You can multi task by cutting the ports and flow testing at the same time.
Honda, Toyota, Cosworth, Ilmor etc. would differ on this opinion.So, the hand ported head, .. that has been tested on the dyno & race track, ..
has to come first, .. and as far as I'm concerned always will.
I would forget about Geomagic or any reverse Engineering Software, I have free access to the best Imageware, for high-end surfacing ; (Ford uses it for scanning and surfacing clay models for car bodies). Even with free access, I never use it. It just isn't a productive step for this type of work. UGNX has very good anaysis tools that show the deviation of points to a surface. there is really no advantage to having junk surfaces in the way.If you have a work flow that uses Geomagic, into Pro-e as an example, ..
you've got over $60K in software plus $3-4K in yearly maintenance, ..
VERY cost prohibitive for a small shop.
Just get a good probe and digitizing software, or maybe a low end coordinate measuring device. Learn to model parametric ports, cut the shape into some port boxes until you get something you like. Cut that into some heads. Make some adjustments if it doesn't fit the casting, grind what you can't reach.
I think you may want to consider removing the hand ported port from the first stages of port design, over time you will find it to be easier and less time with a die grinder in your hands. After reading about the suffering the die grinder causes, I would think you all would be walking over hot coals to escape the die grinder time.
After a few ports have been done the way I am telling you, you will wonder why you ever would start a protoype with a die grinder. When surface modeling becomes as familiar to you as grinding your perspective on this will change.
Yep, and if you market the value of CAD developed ports people will be calling for that and you will sell some heads that you might not otherwise sell.Most customers ask if I CnC heads, .. they don't know why they ask other then
they read in magazines it's the best thing.
yes, and you will make better heads if you develop in CAD.The bottom line is does it make power & win races, and most customers
don't care how I do that, .. the technology being sold is the cylinder head
development not the tools being used to produce it.
Yes that is true for most but some are developing ports and will have advantage from doing so. You can either be one that is or isn't when that method becomes mainstream, catching up is eaier than leading from a technical perspective but it isn't from a business and marketing perspective.The truth is in today's current cylinder head development a CnC machine
is nothing more then a big copy machine. If a bad port design is scanned
and duplicated on CnC, there's not magic HP button on the CnC machine,
all it'll do is duplicate the same bad port design.
When I get back, I will try to make a series of images and maybe some video of port modeling in CAD. I think the unkown aspect of this is making this difficult to accept. Seeing an exact step by step example should clear that up.
too much
Too much subject matter for one thread and there are things I've read that I'd like to counter but I don't want to spend all day here.
Let me just ask a few questions of John.
1st off, you propose a clean slate parametric port model. That much has been stated and is clear. The question I have is this, how do you know that your concept port is better than say, one of Curtis' hand ports?
Just because a part is modeled in a CAD/CAM system doesn't guarantee anything. Can you analyze the port? Yes. Can you claim information from the CAD system in regard to section area, volume and the like? Yes. Can you do much of the same thing with a rubber port core model? Yes you can and I'm sure that many guys here have.
What I'm suggesting is this, and it doesn't necessarily imply that CAD/CAM can't be used:
Take the developed port and have a digitized scan made -or do it in-house-. Create and analyze those surfaces in a dedicated surfacing package. Many surfacing packages ARE parametric and there are a number that will integrate with mid-level modeling packages like SolidWorks.
If you are familiar with surfaces then you also know that once the surface makes it into a CAM package, that surface can be manipulated, deleted and recreated with surfacing tools. Nothing is set in stone.
I've been working with surface models since well before ProE ever coined the word "parametric" and I've hardly ever received a file that didn't have at least one error in it. That's just part of the job. If all I had to work with were native CAD/CAM files it would be nice -although I do have some issues with the "CAM" side of NX & ProManufacturer to name just a couple-, I can create much more efficient high speed toolpathing in PowerMill thanks. Not to mention, I can create watertight PARAMETRIC surfaces in GW3d for about a grand.
If you're clever, you could create a port in something like SolidWorks. The most logical means of creating a port is to create various cross sections along a spline curve and use loft to create your core shape. IMO class A surfaces aren't really necessary for creating this.
Below are two examples:
-They aren't real ports, just three different lofted "sections" and a spline curve driver-.
http://img28.echo.cx/my.php?image=port7cx.png
http://img234.echo.cx/my.php?image=port26rc.png
You could setup relations, adjust control points along the spline and alter section areas with equations in just a few minutes. It took me about a minute to increase the port height from the opening cross section to the minimum cross section.
On the manufacturing side of this. I do not see where building a complete "theoretically optimal" intake port is a smart move. If you want to crank out prototypes, then I would work from a master port, make corrections and slight alterations within a CAD/CAM system then either use FDM (Fused Depositioning Modeling) to make a useable thermoplastic port or build one on a 3D printer -powdermetal type-. Either method would probably provide you with a dozen revisions in a days time. You could build them all at the same time with a large enough printer.
If this is the type of work you're going to be doing on a regular basis then the ROI is there. Much less costly then any of the other methods I can think of at this time.
Your market will come calling based on your successes, not your methodologies. The goal here is building the fastest/least expensive prototype possible. Then ironing it out for another revision. The flowbench will remain your best friend.
edit-
Curtis,
You should take a serious look at a 3D printer. Here's a video I came across when looking for a link.
mms://speedstream.netro.ca/zcorp/PTvideo.wmv
http://www.zcorp.com/
Chuck
Let me just ask a few questions of John.
1st off, you propose a clean slate parametric port model. That much has been stated and is clear. The question I have is this, how do you know that your concept port is better than say, one of Curtis' hand ports?
Just because a part is modeled in a CAD/CAM system doesn't guarantee anything. Can you analyze the port? Yes. Can you claim information from the CAD system in regard to section area, volume and the like? Yes. Can you do much of the same thing with a rubber port core model? Yes you can and I'm sure that many guys here have.
What I'm suggesting is this, and it doesn't necessarily imply that CAD/CAM can't be used:
Take the developed port and have a digitized scan made -or do it in-house-. Create and analyze those surfaces in a dedicated surfacing package. Many surfacing packages ARE parametric and there are a number that will integrate with mid-level modeling packages like SolidWorks.
If you are familiar with surfaces then you also know that once the surface makes it into a CAM package, that surface can be manipulated, deleted and recreated with surfacing tools. Nothing is set in stone.
I've been working with surface models since well before ProE ever coined the word "parametric" and I've hardly ever received a file that didn't have at least one error in it. That's just part of the job. If all I had to work with were native CAD/CAM files it would be nice -although I do have some issues with the "CAM" side of NX & ProManufacturer to name just a couple-, I can create much more efficient high speed toolpathing in PowerMill thanks. Not to mention, I can create watertight PARAMETRIC surfaces in GW3d for about a grand.
If you're clever, you could create a port in something like SolidWorks. The most logical means of creating a port is to create various cross sections along a spline curve and use loft to create your core shape. IMO class A surfaces aren't really necessary for creating this.
Below are two examples:
-They aren't real ports, just three different lofted "sections" and a spline curve driver-.
http://img28.echo.cx/my.php?image=port7cx.png
http://img234.echo.cx/my.php?image=port26rc.png
You could setup relations, adjust control points along the spline and alter section areas with equations in just a few minutes. It took me about a minute to increase the port height from the opening cross section to the minimum cross section.
On the manufacturing side of this. I do not see where building a complete "theoretically optimal" intake port is a smart move. If you want to crank out prototypes, then I would work from a master port, make corrections and slight alterations within a CAD/CAM system then either use FDM (Fused Depositioning Modeling) to make a useable thermoplastic port or build one on a 3D printer -powdermetal type-. Either method would probably provide you with a dozen revisions in a days time. You could build them all at the same time with a large enough printer.
If this is the type of work you're going to be doing on a regular basis then the ROI is there. Much less costly then any of the other methods I can think of at this time.
Your market will come calling based on your successes, not your methodologies. The goal here is building the fastest/least expensive prototype possible. Then ironing it out for another revision. The flowbench will remain your best friend.
edit-
Curtis,
You should take a serious look at a 3D printer. Here's a video I came across when looking for a link.
mms://speedstream.netro.ca/zcorp/PTvideo.wmv
http://www.zcorp.com/
Chuck
Yes, while Geomagic does create a quiltwork of surfaces, it uses a very powerful smoothing algorithm to get rid of surface-to-surface blend problems. The software is not as useless as you are being led to believe.cboggs wrote:Jon,
Doesn't GeoMagic let you do this?#1 You absolutely must have parametrics so that you can adjust dimensions of your design and all the surfaces will still fit together instead of needing to recreate new surfaces.
You can scan on a machine but it will be a very slow process. I personally would rather be making chips with a machine instead of delegating it's time to scanning but that may be just the thing for such a low volume?Harbinger,
You bring up several good points.
first, sending stuff out to get scanned & CnC ported, like I do now, ..
has several draw backs, the big one is costs, ..
To scan and prep the files most people want from $1K to $3K, ..
then $100+ an hour machine time to cut, .. I can't resell the work and
make a profit.
If I'm cutting 20 heads, yes it's worth it, .. if I'm doing two sets, ..
it's no where near worth the investment.
I'm a custom shop, .. doing small runs of stuff and need to do
the CnC in house for that reason. Fast turn around is another, ..
and keeping some designs all "in-house"
I'm with Shawn that what I really want is a simple solution to
increase my productivity, .. to CnC bulk material and hand finish
the heads.
So from that stand point, .. using Geomagic & Surfcam, .. and a 3 axis machine
with a fixture or even a 4 axis machine and position cutting sounds
like a very good, .. and the "simplest" plan.
My question is this, .. Jon, .. Harbinger, .. couldn't both work flows be
used depending on the need? Geomagic & surfcam for the hand port
& copy work, .. and Pro-E for the deeper development work?
Now about the scanning part, .. can I digitize in the spindle with a
touch probe?? Fero arm, .. ???
Oh, .. Harbinger, .. what's your first name? pet peeve of mine, ..
I'd like to use your name to properly address you if you don't mind.
Curtis
Every one's manufacturing situation is different. That's why grade A, B and C software packages can survive and remain profitable in this industry. Not everyone needs, or can afford, a Catia.
Back to scanning-
Offline scanning like Renishaw Cyclone would be a fast and sensible choice. CMM time is just too costly for this type of duty.
If you will take the time to analyze your business situation, I'm sure you'll come up with the right processes to achieve your goal.
What I would be interested in is building a modular intake manifold, whcih could easily exchange different runners, plenums and the like. A rapid prototyping system that will build a usable model is, in my opinion, a good candidate for all of this development.
Chuck
If you ever make it down to Texas let me know John. I will show you how to do this successfully. We've been doing it for some time now.SchmidtMotorWorks wrote: About four years ago I was tasked with evaluating variuos Reverse Engineering software. I spent about 6 months using about 8 different softwares to model a hood and fender from scanned points or facets whatever the software required. None of them produced direcltly usable surfaces even with the top AEs from each company showing how to use it. At best I had a good reference to use for a parametric model.
Believe it or not but many people still develop products the old fashioned way, by employing a model maker to build a prototype. I've scanned a number of them over the years and have to this day, never had a problem recreating the surfaces and/or cutting the molds. One company we make bottle molds for used hand shaped models up until just last year. I was cutting molds for them almost 6 years prior to that from master models they'd supplied.
Next time you're out trying on software give Mastercam's MoldPlus a look. It does exactly what you are describing -mold cavity & parting line creation- effortlessly. Better yet, Mastercam is relatively inexpensive as far as CAM systems go and has better toolpathing options than the top tier. Does an excellent job with electrode creation too.Another project I worked on a couple of years ago was testing a mold cavity splitting software on about 300 different plastic parts from all over the world everyting from dashboards to hairdryers you name it modeled in every software I ever heard of and more. One thing that I remember about that test is that we were looking for examples of parts that were produced from Reverse Engineering software to use in the testing. Not one of the mold shops we had as customers had made a mold for a part that was designed for production with a Reverse Eningineering software.
If it could really do what they claim, there would be more products designed with it.
Chuck
-
learner
Chuck, the link to the 3D printer is awesome.
I'm a beginning user of Solid Works. I have no knowledge or experience of what you guys are talking about for the most part. I have played with modeling port molds and currently i'm unable to finish it correctly. I'm assuming it's due to my lack of skill. I do know that modeling a wedge port isn't as easy as lofting a port such as a hemi head or the one that Chuck showed. At least not for me anyway. Here is what i've been able to do with an intake mold.
Could someone tell me what i'm doing wrong or provide a couple of suggestions.
Thanks
How about this?
http://pg.photos.yahoo.com/ph/autorodmaker/my_photos
I'm a beginning user of Solid Works. I have no knowledge or experience of what you guys are talking about for the most part. I have played with modeling port molds and currently i'm unable to finish it correctly. I'm assuming it's due to my lack of skill. I do know that modeling a wedge port isn't as easy as lofting a port such as a hemi head or the one that Chuck showed. At least not for me anyway. Here is what i've been able to do with an intake mold.
Could someone tell me what i'm doing wrong or provide a couple of suggestions.
Thanks
How about this?
http://pg.photos.yahoo.com/ph/autorodmaker/my_photos
Last edited by learner on Mon Jun 13, 2005 9:23 pm, edited 3 times in total.
It worked for me.
I've been working on this too.
I created a path line to loft along and used additional planes to make cross sectional area sketches to use as profile segments for my loft. I've got a pretty sucessful exhaust port so far. I'll keep working on it to see what I can come up with. This thread has prompted many questions and ideas for me.
Maybe Don can do a CD for this subject or a book or something? Seems like a hot topic with a number of solutions. But this is definately the future of head porting as far as I can see.
Sean
I've been working on this too.
I created a path line to loft along and used additional planes to make cross sectional area sketches to use as profile segments for my loft. I've got a pretty sucessful exhaust port so far. I'll keep working on it to see what I can come up with. This thread has prompted many questions and ideas for me.
Maybe Don can do a CD for this subject or a book or something? Seems like a hot topic with a number of solutions. But this is definately the future of head porting as far as I can see.
Sean
-
learner
SWB, that is a different link than what shawn tried, i had to change it.
I have messed with lofting also but for an accurate port you would have to make a new plane every .005" or .010" or so and i had problems doing that around the short turn because the planes would intersect each other and would let me create new ones because of that. I have no doubt there is a way to do it, i just currently don't know how. I can see it working on an exhaust port due to the difference in shape.
I have messed with lofting also but for an accurate port you would have to make a new plane every .005" or .010" or so and i had problems doing that around the short turn because the planes would intersect each other and would let me create new ones because of that. I have no doubt there is a way to do it, i just currently don't know how. I can see it working on an exhaust port due to the difference in shape.