straightness or flatness datum

[duk748]

hello - i have a leveling plate that will be made from a36 h.r.s. that is to be used to level 3 pads for a table - the plate is only 2mm thick - not machined either side - i want to hold some type of datum so that the plate can be made straight or flat - shims will be used on top of this plate to achieve the correct level after it is installed - any help would be greatly appreciated - thank you - plate is roughly 960mm x 860mm

 

[KENAT]

I think your terminology may be a bit mixed up.

Flatness doesn't reference a datum, it's essentially variation of the surface to itself.

For you application maybe put a flatness tolerance on one side of the plate and make that surface say Datum A. Then make the opposite side of the plate parallell to Datum A within required amount.

However, I'm not sure I fully understood your question.

KENAT,

Have you reminded yourself of faq731-376 recently, or taken a look at posting policies:

http://eng-tips.com/market.cfm?

 

[duk748]

hello again & thank you - yes after taking the time to look in my gd&t book i realized that i was mistaken about straightness - flatness is the way to go exactly as you have stated here & yes you did understand what i was saying - i should have thought about it a little more before posting - sorry - thank you again

 

[NomLaser]

Do you really need the plate to be flat if you are using leveling feet anyway?

Flatness is a really difficult and time consuming tolerance to measure, where straightness is an offset plane 1/2 the tolerance in either direction from the nominal dimension.

http://www.tec-ease.com/tips/july-99.htm

 

[jerry1423]

Straightness is taking measurement readings along a staight line (one dimensional), and the deviation of those readings.
Flatness is is taking measurement readings over an entire plane (two dimensional), and the deviation of those readings.
In your case you probably want to use flatness even though it takes more time to get good readings.
I am not sure if I gave you what you are looking for but I hope it helps.

 

[drawoh]

duk748,

Are you talking about specifying datums and tolerances on your drawings, or are you talking about some sort of fixture?

I do a lot of drawings of aluminium parts 2mm[ ]thick. Generally, I attach a note to the face datum stating that it applies when the part is clamped to a flat reference surface. Even a steel 2mm[ ]plate is going to bend under gravity, and possibly under the load of your dial gauge, depending on your required flatness.

JHG

 

[duk748]

hello again & thank you for all the replies - this is why i come to this board - sometimes the more experienced engineers/designers can really provide a nice service to those of us who do not have adaquate inside resources - i thank you again & everyone have a wonderful healthy & prosperous holiday season

 

[NomLaser]

"Straightness is taking measurement readings along a staight line (one dimensional), and the deviation of those readings.
Flatness is is taking measurement readings over an entire plane (two dimensional), and the deviation of those readings.
In your case you probably want to use flatness even though it takes more time to get good readings.
I am not sure if I gave you what you are looking for but I hope it helps."

IMO, if he is using shims to achieve the correct height anyway, then there really is no reason to add more cost to the part than what is needed. Yes flatness would be a better measurement if he were mounting directly to the plate. Just my opinion based on what we have been told.

 

[KENAT]

He never said the flatness would be tight DEddie. Certainly it doesn't seem to make much sense to go to a lot of effort to have a very flat plate if you're using shims on each foot as required anyway. However, it may not be unreasonable to have some flatness to preclude gross bowing etc of the plate.

KENAT,

Have you reminded yourself of faq731-376 recently, or taken a look at posting policies:

http://eng-tips.com/market.cfm?

 

[PaulJackson]

Jerry1423

Straightness is taking measurement readings along a staight line (one dimensional), and the deviation of those readings.

Not if the straightness is beneath the size callout... in that case it pertains to the axis or median plane... and if the tolerance is modified (M) or (L) it establishes a virtual condition limit.

See DEddie's tec-ease link where Don Day explains that.

Paul