
Dimentions and drawings.
#1
Posted 19 April 2011 - 07:19
I work with "engineering" But Actually more like design as i rarely calculate the stuff i make. i usually just use the Ansys in Inventor.
And in my opinion the firm i work in do a lot of strange things. Such as putting on every single hole dim when they could just use a symmetry line and put on dimention on just one side as the other side is equal. Problem is that i don`t have a document or a good source for that
to be normal drawing practise. I just think i heard it during a course i went on. To me it should do with one hole dimention if all the others are equal and a symmetry line if the plate is mirrored across the sentre.
They also put up the plate thickness either as text on it or simply show the side of the plate with the thickness dimentioned. I think it should be enough with the title field. where the outer dims are.
Silly place.. makes me hear monthy pyton songs in my head..
Advertisement
#2
Posted 19 April 2011 - 08:40
#4
Posted 19 April 2011 - 12:24
1) The only number that matters on the drawing is the authors phone number
and 2) If in doubt on dimensions leave them all off and just show the scale of the drawing.
Both presumably now irrelevant with CAD/CAM ?
#5
Posted 19 April 2011 - 14:28
. Work with Annotations
. Hatches, Fills, and Wipeouts
. Tables
. Dimensions and Tolerances.
In Architecture, Landscape Architecture and some forms of Engineering it is appropriate to use the word 'Typical' when referring to a repetative graphic element. Too many dimension lines make drawings very difficult to read.
In any event, labeling is in my experience an art...just make sure everything adds up! The over-all and all the incrimental dimensions between need to reconcile...there may be a few spans that are not dimensioned on purpose to help un-clutter the drawing. In those cases the un-dimensioned span must be verifiable based on all the other dimensions. For example, if we divide a square into 4 eqaul quadrants we can dimension two over-all lengths and one of the quadrants leaving the rest to be verified by those two combinations. Obviously a square or cube is the simplest one dimension shape or form we can use.
Re no scale...I wouldn't allow anyone to scale a drawing directly. If that becomes necessary then the drawing is incomplete...in my opinion.
Edited by meb58, 19 April 2011 - 14:37.
#6
Posted 19 April 2011 - 16:20
Two comments always found on 'traditional' engineering drawings - DO NOT SCALE, and IF IN DOUBT ASK!Re no scale...I wouldn't allow anyone to scale a drawing directly. If that becomes necessary then the drawing is incomplete...in my opinion.
#7
Posted 19 April 2011 - 18:11
#8
Posted 19 April 2011 - 18:54
I took a course on Geometric Dimensioning and Tolerancing based on the ASME Y14.5M-1994 standard. I found it to be difficult to understand at first - like reading hieroglyphics but by the end of it was a big fan. It leaves nothing to guesstimate or "industry standard" or any other dubiously undefined conclusion.
As a normal practice, we dimensioned everything that mattered, say the overall length, the width, depth and height of the pockets, the location of the pockets but not the space between the pockets as that is ultimately controlled by the location and length dims and tolerances - and we don't care about that space. Alternatively, we may dimension one pocket's location based on another - in which case the space between them is of concern, but the overall placement of the 2nd pocket relative to whole body is not.
#9
Posted 19 April 2011 - 21:54
The other bugbear was clients who insisted on designing things they weren't qualified to specify - Vacuum formings with no draft angles and constant section thicknesses; tolerances specified to metric 3 decimal places. I once received a drawing for an ABS moulding 600mm long +/- 0.001mm - I rang them up and asked "What day of the year, and would they care to specify the temperature and humidity at which these tolerances applied?" They hadn't even bothered to look up the coefficient of linear expansion for any plastic, let alone one that is also hygroscopic...
Edited by Bloggsworth, 19 April 2011 - 22:06.
#10
Posted 19 April 2011 - 22:17
Your drawing office may also have a manual of style.
Now the chances are that the above may as well have been written in Martian for all its applicability to a typical CAD shop. So Mats, can you describe roughly what happens between you having a decent 3D model of the part you want made, and the arrival of the first part made off the drawing?
Edited by Greg Locock, 20 April 2011 - 00:26.
#11
Posted 20 April 2011 - 12:12
#12
Posted 20 April 2011 - 13:34
What happens from when i have a 3D figure to its made?
First step is to make sure i have a ladder layout as in the main figure first, the sub assemblyes next then parts under the relevant assemblies. I do this in order to make it easy to separate the drawings and sending it to the different divitions.
Then its a endless check of dimentions and going into the 3D drawing again and fixing the figures so that the figues is a round value. 50mmblank etc. they usually are from the start but sometimes not.
then its checking that there is no dimentions lacking..
Then there is making sure its not more dimentions than needed. I go and use the symmetry line alot to symbolise that its equal on both sides. it leaves space on the other side to dimentions that did get the clean space i had on the other side etc.
there is actually a lot of work before you get to paper that is relevant. because i worked as a tool maker a little bit in my apprentice time and had a unusually high amount of machining/tool making for a industrial mechanic. I also did a lot of repair and my "mentor" was very into smart designs all the way from big machines to hand tools, so i got focus on it early.. As a mechanic you notice when a screw is behind, down, under a gearbox.. when it did not need to be..
anyway lost track there.. the key is to think machine park and easiest way to make it work.
I usually try to make as much as possible from the same thickness. like 10mm plates. we don`t need one 2x2m plate in 10mm and one in 15 and one in 30mm and so on.
KISS
keep It Stupid Simple.
or
Keep It simple Stupid.
I only make complex machining pieces if i can save time in the production or improve the accuracy or make a improvement to health and safety and so on.
For me its a balance akt.. how far can i go? i want to make it as easy to use and as fast to use as possible.
What i also try to do is to give one part multiple functions. say the same piece can disassemble and mount by simply turning it around.
Usually by the time i have tried to do all this and it leaves me i get a bit "blind" and a few dimentions gets forgotten. then the machine man asks for it, i fix and prints out a new one.
Tolerances. i keep it as rought as possible.. on a haste job i once made a 500+100mm tolerance. because it did not matter.. just cut the ****er with a saw and send it over there, its not relevant anyway. I wish it was a "this is rough so don`t waste time on it" symbol..
I only make a fine tolerance if its recuired... No demands but i know the angle of the part that goes there is important = down to 0.1+-mm in the mill. i consider that a fine tolerance to linkage on trains...
I once made a press tool that had as low as 0.00 to 0.05 gap allowed.. i had to do it because it had some fine alignment tolerances inbetween the parts that where to be pressed down aligned to each others. you can`t see a gap on the parts after mounting eventho the parts had a roughter tolerance on their thickness than what would have been required for to maintain the fine tolerance on the alignment.
It does it more aligned than factory mountings and the tool works like a dream being faster and easier and taking less space than the previous stuff.. it just goes to show how stupid fine tolerances sometimes can be..
Materials are cheap and if its not going to be lifted by hand. just keep it beefy.
#13
Posted 20 April 2011 - 13:39
I was looking for alternative ways to show f.ex that the hole is countersunk and so on. Just different ways of doing things so i know my options. I find it anyoing that i have to show the side of a plate to show how deep the countersunking goes. and so on.
on round things i wish it was ok or within a standard to give the hole dim and then type 2x45* (insert degree symbol instread) to show that the hole gets a chamfer or a rounding if thats the right word.
#14
Posted 20 April 2011 - 14:36
A chamfer is shown in both plan and side view...both clarify one another. With drawings and annotations I take the perspective that it is not enough to be understood, one must not be misunderstood.
EDIT: I am participating in a CAD Webinar as we speak - er write. One of AutoCads newest features helps Annotation productivity and accuracy. You choose your profession are area of speciality and the the convention is automatically set up without you having to select it or wrry about it - applies the attribute in the correct layer and in the correct color. So, although this new productivity tool does make us a bit dumber, it can free up a lot of time over the course of preparing a large document/drawing.
In addition, you can click and drag annotative icons from this library right onto your drawing and they will be, as I wrote above, on the corect layer, correct line weight...color etc.
FYI
Edited by meb58, 20 April 2011 - 17:08.
#15
Posted 17 October 2011 - 11:37
I believe i was tought about this simplifying idea in my course. But im struggling to find some hard facts about it.
Its used to show that the piece or assemly is equal on both sides so that you only need to put for instance weld symbols on one side. That way you get free space to pos numbers and dimentions on the other side. It cleans up the drawing quite a bit.
Fun fact: the drawings on my work place does not have a reference to a dimention standard. I think i need to suggest that so i get a course.
Edited by MatsNorway, 17 October 2011 - 11:37.
#16
Posted 18 October 2011 - 07:47
ISO128 - 1982 and NS 1404 has it.
Symmetry lines are marked with two parallel lines on the end of the symmetry line.
Likes this(imagine the parallel lines to be straight)
// --- - --- - --- - --- - ---//
It does not say anything about assemblies tho.