Allowable stress for structural members - t/R ratio

[Pato12]

Hi Guys,,

Just wanted to confirm with other engineers here is the R in t/R, the outer diameter of the structure member in question?

I copied the notes at the bottom of table 5.3, its says the following;

"The variables in the compressive stress equations are defined as follows:
l= unbraced length of the column, in in.; r= corresponding least radius of gyration of the column, in in.; t= thickness of the tubular
column, in in.; Y= unity (1.0) for values of t/R equal to or greater than 0.015; Y= (2/3)[100(t/r)] {2–(2/3)[100)t/r)]} for values of t/R less than 0.15."

 

[IFRs]

Which revision of which code?

 

[Pato12]

Oh dam sorry I keep thinking this part of the forum is just for api 620.

Im using API 620 - latest one with 2012 changes.

Ive looked at older versions and they have the same table and definitions.


I also curious is there a more dated version of this standard because I have noticed some typos here and there and even sometimes missing math operators like division and multiplication in formulas. For example I strongly believe the tubular axial loading allowable compressive stress whould be l^2/18000*r^2, not l^2*18000*r^2. You get a very small stress regardless of l and r values otherwise.

cheers,


pato

 

[JStephen]

One of the assumptions used in much of API-620 is that shell thickness is small compared to overall dimensions. With that in mind, R would generally be taken as the nominal radius. So for a 50 ft diameter tank, R would be 300 inches, regardless of whether the tank was 50 feet ID or 50 feet OD or 50 feet to plate centerline. Plugging some numbers into the stress formulas and comparing gives around 0.6% maximum deviation figuring inside radius vs outside radius, much less for smaller t/R.

On the compressive stress calculation, compare the line for tubes to the line above it for structural members- there is in fact a divide sign that is missing. Also, it's not obvious, but the r is part of the denominator- it is usually written (L/r)^2/18000, rather than split up as shown.

 

[Pato12]

So just wanted to confirm with what you were saying and no doubt I just confused to hell just now but here goes.

When finding Y for the compression equations for structural members, the t/R ratio that governs Y is,

t(thickness of structural member) / R (distance from compression member's centreline to the sidewall)

is that right?

and second,

is the equation of the compression member the following,

18000(Y + ( 1 + (l^2/(18000r^2)) ) )

and finally,
the equation of Y for t/R less than 0.015(which by the way says 0.15, clear typo there) is the following,
Y = (2/3)[100(t/r)]*[2 –((2/3)(100(t/r))], where
t=member thickness
r=radius of gyration


Are my formulas and what each variable definition right?

 

[Soln]

The formula is actually:

18000Y/(1+(L/r)^2(18000))

And,

You're right, there's a typo but, I think the correct number is 0.015.

Using 0.15 would be very conservative for a structural column, which I think API 620, Table 5.3 is intended for. Considering an 18 inch diameter column, the thickness would be 1.35 inch before the user would get away from the reducing coefficient Y.

Can anyone source this portion of 620? I'd also like to know the source of the formulas in API 620, Section 5.5.4, and API 650 E.6.2.2.3? These allowable compression stress equations seem related.

 

[JStephen]

It looks like the 18,000-equation would be based on the Rankine-Gordon formula, with adjustments to the constants and FOS.