How to create this cam in NX8.5: Good challenge for NX expert 4

[PSI-CAD]

Hi,

I would like to obtain the same result than PRO/E in NX8.5

There is a function evalgraph in PRO/E

The graph is below:

I have created the GRAPH with a sketch and I didn't find the way to use it in a sweep or var sweep

Find herewith the NX8.5 part file

Thanks in advance

Regards
Didier Psaltopoulos

http://www.psi-cad.com

 

[NutAce]

Hi,

I had to scale the length of your sketch to match the circumference of your "pipe".
I only did this for the length so you probable need to do this for the rest of the sketch geometry as well, but it will give you a general idea on one possible solution.

Ronald van den Broek
Senior Application Engineer
Winterthur Gas & Diesel Ltd
NX8.5.3 / TC9.1.2
HPZ420 Intel(R) Xeon(R) CPU E5-1620 0 @ 3.60GHz, 32 Gb Win7 64B
Nvidea Quadro4000 2048MB DDR5

HP Zbook15
Intel(R) Core(TM) i7-4800MQ
CPU @ 2.70 GHz Win7 64b
Nvidia K1100M 2048 MB DDR5

 

[mmauldin]

My take on this, NX7.5:

 

[PSI-CAD]

Hi,

I had already try this

Thanks for you help but wrap after scale is not the good solution because it's not enough to scale the length

Have a look at my approximation

Regards
Didier Psaltopoulos

http://www.psi-cad.com

 

[JohnRBaker]

Here's my take on this. Note that while it appears that I also scaled the sketch to fit the 'length' of the outside of the pipe, I only scaled it along it's length by adjusting the linear sketch dimensions so that they stayed proportional keeping the height dimensions constant.

John R. Baker, P.E.
Product 'Evangelist'
Product Engineering Software
Siemens PLM Software Inc.
Digital Factory
Cypress, CA
Siemens PLM:
UG/NX Museum:

To an Engineer, the glass is twice as big as it needs to be.

 

[Chigishev]

Two law curve, one for inner radius, one for outer.
X=rcost
Y=rsint
Z=law curve from your scketch, no need to scale.

 

[PSI-CAD]

Hi John,

Did you add your part to my assembly to compare ?

The result is not the same

If you look a bit further, you will understand that we need to reproduce a cam operation. The length 360 is in fact 360°

Hi Chigishev,

It seems to be the solution, but I didn't succeed: have a look to my part

Thanks in advance

Regards
Didier Psaltopoulos

http://www.psi-cad.com

 

[t300master]

Hello,

you'll missed the 360 degree as a factor, see attached example.

Best Regards
Meinolf

 

[PSI-CAD]

Hi Meinolf

You are the winner

Thanks a lot

Regards
Didier Psaltopoulos

http://www.psi-cad.com

 

[Chigishev]

And a lot of different solutions. You can build ruled surf with two curves. I build swept. Not so important...

 

[PSI-CAD]

Hi Chigishev

I have forgoten to congratulate you

In fact, you are the winner

Thanks a lot

Regards
Didier Psaltopoulos

http://www.psi-cad.com

 

[Chigishev]

Thanks!

 

[cowski]

When designing cams it is common to specify the "dwell time" (time the follower is in a constant position) and "rise time" (time the follower is moving up or down). When the angular speed of the cam is known, you can talk in terms of time; however, it is common to use angles for the dwell time and rise time as this is independent of the speed of the cam. The sketch in the OP is 360 units long; therefore each single unit of length in the sketch represents one degree of cam travel. The linear dimensions are the ones that are important (they should be driving dimensions); they will govern the angular dwell and rise times of the resulting cam profile. The cam is divided into three identical 120° sections; each section starts with 52.2° of rise followed by 43.2° of dwell. The remainder of the 120° section falls back to the original position. The angular dimensions in the sketch are unimportant. By reworking the sketch and using the techniques shown by others above, you could allow easy user input of "rise time", "dwell time", and rise distance. This would allow easy, direct changes to the cam profile.

www.nxjournaling.com

 

[Ehaviv]

here a revision of t300master part
you can play with the parameter n
to change how many times the shape Repeated
crrently n=3

 

[Ehaviv]

sorry here the improved (correct) part

 

[Ehaviv]

add spine to swept function