【转帖】dimensioning of profiled surfaces

yang686526

dimensioning of profiled surfaces
i'm relatively new to the asme y14.5 and have been using solidworks for a year. we're outsourcing some investment cast aluminum parts which require secondary machining. while we can assign dimensions and tolerances for the machined features (holes) on the drawing, we are a bit puzzled as to how best to locate and dimension several as-cast features (eg. ribbed pockets with draft and radiused edges) which do not require any machining. the features in question are primarily for weight reduction. how should we do this?
it's not yet clear if the foundry can reference the solid model directly, but it's encouraging to see that supplying a drawing plus model to the manufacturer appears to be common practice among the posters here.
the as-cast dimensions are generally not toleranced. the manufacturers standard tolerances will be acceptable.
any comments or suggestions are most welcome.
many thanks!
tom
eng-tips forums is member supported.
the surfaces/features that are not machined later should have dimensions with tolerances. all others can have dims using a tol block (loose tol) and/or general notes to cover radii, edges, etc. a big help could be to send the foundry a model and dwg with dims and ask them to mark-up the dwg for you. sometimes a foundry will do this for you.
chris
sr. mechanical designer, cad
solidworks 05 sp2.0 / pdmworks 05
chris,
the problem we have is that the surfaces we'd like to dimension are not square. beats me how to locate the sloping surface of a drafted pocket with radiused edges, short of supressing the radiused edges in the drawing and covering the edge radius with a note. this still presents problems during inspection.
should the foundry be milling the pattern via cnc, then it will be easy enough for them to generate toolpath directly from the solid model.
conformance with asme y14.5 is not required. this project doesn't warrant the time and effort to master this rather obtuse spec.
regards,
tom
can you show us a pic of your part? i can have a better understanding of what it is.
also, quick note ... y14.5 should be req'd for all parts if you want consistant parts in your product. not many master the spec. it is confusing to most, but is worth the time.
chris
sr. mechanical designer, cad
solidworks 05 sp2.0 / pdmworks 05
the part is proprietary, so i prefer not to post a picture. perhaps we can discuss a hypothetical case.
visualize a 6"w x 12"l x 2.5"h block of aluminum. there is a 3"w x 6"l x 1" deep pocket at the center of each of the largest sides. each pocket has 2 x 0.25" thk ribs connecting diagonal corners of the pocket. all internal side faces of the pocket and ribs have 3 deg draft. all edges are radiused 0.10". the sides of the block are orthogonal and can be used as datums.
regards,
tom
it seems you wish to control the non-critical, as-cast features of your casting.  i've done a similar thing by dimensioning to the center line of the rib(s) and defining the typical rib (usually a section detail) as either the nominal bottom width minus draft or nominal top width plus draft.  fillets and rounds were dim'd as "typ".  as to the edges of the pockets themselves, i've dim'd to the theoretical sharps and added the plus/minus draft symbol as required.
we defined a small circle with either a plus or minus sign in it as a symbol to add to dimensions and defined what they meant in the general notes.  for example, minus draft (as for a hole/pocket) was defined as; "dimension is at largest point of feature due to draft".
if there could be any confusion, add a note that pocket (etc) edge dims are to theoretical sharps.  "when in doubt, spell it out."
providing the model to the foundry is very helpful.  if they can't read it directly (same software) provide them a 3d solid they can use (iges, step, parasolid, etc.).  if they can't read 3d models, at the very least, create section views of all important features on the drawing(s).
redpen
tom,
i worked with helmet a div of alcoa in canada on a c355 al pump housing.  i chose three cast datums and used a surface profile of .060 which equates to +/-.030.  this was an extremely complexed cast part and they requested a solid model to work the tooling.  it had a lot of internal pathways that were out of plane.  just as a side note 3 degrees of draft for investment casting is a lot....i specified 1.5 degrees without any problems....that's the beauty of investment casting.  my advice is to talk with your casting vendor and see what they feel comfortable with.
best regards,
heckler
sr. mechanical engineer
sw2005 sp 2.0 & pro/e 2001
dell precision 370
p4 3.6 ghz, 1gb ram
xp pro sp2.0
nivida quadro fx 1400
      o
  _`\(,_
(_)/ (_)
do you trust your intuition or go with the flow?
why not just put a general surface tolerance on it?
use machined faces as datum surfaces and a hole as positional, you know have you’re a,b and c datums, everything else can have a surface tolerance from these.
with a cmm and half decent software the model is the master, you can of course specify specific areas you wish to check.
i have no idea if this falls within american standards but it is certainly how it is done in the european automotive industry. tie up anything that matters squareness, parallelism, bore diameters, positions and the like and just give a general, surface or trim tolerance for everything else, all taken from the model.
redpen,
thank you for your excellent suggestions. sounds like a perfectly workable approach for the as-cast features.
kind regards,
tom
heckler,
yes, the approach taken really depends on the tool manufacturing method, which as yet is not known. the question has been asked and we're awaiting response from the overseas supplier. if they can work wth the solid model directly, so much the better. ;)
the foundry suggested 3 deg of draft and 6mm minimum wall thickness. i expected less draft, but can live with 3 deg. we're not making parts by the millions, so can afford to be flexible.
best regards,
tom
ajack1,
thank you. should the foundry be able to work with the solid model directly and have cmm capability, then the approach you suggest looks like a good way to go.
however, i suspect the foundry may not be so sophisticated, in the metrological sense.
perhaps a case of the blind leading the blind. ;)
best regards,
tom