When i make an harmonic analysis,
how sif take Caesar ?
I think the sif named i from B31.1 for a tee is not the peak stress concentration faktor.
Caesar take the dumb rule 2 times i = CAESAR II*K2 ?

I must assume that by "dumb" rule you mean to say "simple" rule rather than "stupid" rule.

Either way, we don't do it. CAESAR II (CAESAR II) uses the stress intensification factors (SIF's) established by the respective piping code to calculate harmonic stress.

You are correct, the SIF's used in CAESAR II should not be confused with a stress concentration factor used in peak stress calculations.

I have avoided such peak stress analysis directly in CAESAR II. CAESAR II works well for system analysis but it should not be confused with a local analysis tool. Remember that we are working with a simple "centerline" model with no detailed component geometry defined. Certainly, you could use CAESAR II's element forces and moments as loads for a local analysis (read: FEA).

Three other points need mentioning:
1) The harmonic stress calculation used in CAESAR II is dependant on whether or not a number of cycles is specified. If load cycles are not specified, the stress category in output is labeled (HAR). If load cycles are specified, the stress category in output is labeled (FAT).
2) If number of cycles is specified, (FAT), CAESAR II will calculate a "3D Stress Intensity". A fatigue curve should be included in the (static model)input to set the allowable limit. Read up on how to do this. Here's where you can introduce your (dumb?) "safety factor" of two.
3) If no cycles are specified (HAR), then harmonic stresses are calculated using the code equation for occasional stress.

If you don't want to fool around with counting cycles and fatigue curves, keep your alternating stress below limits suggested in API618 or OM-3; that is below about 45MPa.

I think Mr.Ohliger means a "rule of thumb" rather than a stupid rule

Thanks !
I had mean "thumb" and not "dump".
My mistake !
My following conclusion right ? :
Caesar calculate stress amplitude with the SIF
how Caesar me show in result stress output listing. This SIF is not a peakstress SIF.
With this stress "Sa" give Caesar the allowable load Cycles. With the input load cycle
i have then the damage.
Sorry for my bad english

If I can summarize...

For fatigue analysis (for codes other than TD/12), CAESAR II calculates a maximum 3D stress intensity. This is what's used for harmonic stress calculations where load cycles are defined. (I incorrectly stated previously that stresses are shown as (HAR) when load cycles are not defined; that's been changed to (OCC).)

CAESAR II on-line documentation has an awkward expression for that 3D stress intensity. I have a pdf posted on our web site that says it a little better. See 3D S.I. .

Now, for harmonic analysis on CAESAR II, where pressure equals zero this gets real simple. Hoop and radial stress are zero so all you have remaining is longitudinal stress and shear stress. The maximum 3D stress intensity reduces to SQRT(SL^2+4T^2).

This is a stress amplitude as is goes both positive and negative. Compare this to the material fatigue curve that's based on amplitude (rather than peak-to-peak). Since we're probably talking high cycles here, instead of accumulating damage, just be sure the stress amplitude remains below the endurance limit. For that I'll point you back to API 618 or OM-3.

Yes, we are not considering paek stresses.