PROCESS AUTOMATION

26 PROCESS Worldwide 1-2017

MANAGEMENT OPERATION MAINTENANCEDESIGN

OIL/GAS

WATER/WASTEWATER

CHEMICAL FOOD/BEVERAGE

PHARMA/BIOTECH

It was 42 years ago that theAmerican Society of MechanicalEngineers (ASME) published a

standard for temperature measure-ment—PTC (Performance TestCode) 19.3-1974. The provisionsthat it contained for the wake fre-quency calculation were based onJ. W. Murdock’s fundamentals from1959.This standard remained un-

changed for more than 20 years. In1995 there was a serious accidentat the Japanese nuclear powerplant in Monju. It was caused bythe rupture of a thermowell due toturbulence: The vibration directionof the tube was parallel to the flowdirection of the medium in thepipe. Such an in-line resonancewas not considered in the original

PASSING THROUGH

Source:W

ika;©BitsandSplits,©PF-Images/Fotolia.com

;[M]GötzelHorn

Wake frequencycalculation:What haschangedwith the 2016revision of the ASME PTC19.3 TW standard? — Itstands as the benchmark:Those who operate withthermowell strength calcula-tions in accordance with theglobally recognised standardASME PTC 19.3, will protectthemselves against alleventualities.With theadditional mark of“TW-2016”, the standardspecifications are nowavailable in a revisedversion. The revisionprovides some clearer andmore-detailed definitions.

K A I G R A B E N A U E R *

* The author is Product Manager, CoE EuropeElectrical, Temperature Measurement,Wika Alexander Wiegand SE & Co. KG,Klingenberg/Germany.Contact: Phone +49-9372-1320

THE IN-LINERESONANCE

document2483191793277848697.indd 26 27.02.2017 14:54:33

1-2017 PROCESS Worldwide 27

calculation basis from 1974. Thisedition of the standard only con-sidered the more frequently occur-ring vibrations at right angles tothe flow direction.As a consequence of the Monju

incident, the ASME standard from1974 was completely revised andthe result was published in 2010.Its most important improvement incomparison to the original is-sue—apart from a variable Strou-hal number—was the inclusion ofthe in-line resonance (which hadbecome the problem in Monju) inthe calculation. Since then, ASMEPTC 19.3 TW-2010 has been ap-plied successfully worldwide in allindustries whose processes aresubject to high loads such as thosearising from high flow rates. Thisapplies, for example, to companiesin the oil, gas and chemical indus-tries.Now ASME PTC 19.3 TW-

2016—an updated version—isavailable. The amendments con-tained within it relate to the fol-lowing three aspects.

Requirements forpassing through thein-line resonance

In chapter 6-8.5 the require-ments for passing through the in-line resonance with a ratio of theStrouhal or excitation frequency tonatural frequency fs/fn = 0.5 aremore clearly defined.As a result, passing through the

in-line resonance is possible, if theprocess medium is gaseous andthe range of the in-line resonancedoes not fall within the continuousoperation of the plant. As expectedpassing through this in-line reso-nance is the responsibility of theplant owner. Here, the permissibleflexural fatigue stress must not beexceeded during the cumulativenumber of vibrations of 1011 cycles.How this translates into figures isshown in the table. The calcula-tions refer to a solid-machinedthermowell from stainless steelwith a tapered stem, 25/19 mm di-ameter and 6.6 mm bore.Furthermore, the process medi-

um must not adversely affect thefatigue strength of the thermowellmaterial. For the case, which can-not be completely eliminated, of a

thermowell rupture resulting fromin-line resonance, the standard de-mands safety provisions to preventsevere injury to personnel or dam-age to the plant.

Installation in pipeelbows and angledinstallation in pipelines

With the mounting of thermo-wells in pipe elbows, “TW-2016”distinguishes between two funda-mental cases (see Fig. 1): In case1, the thermowell tip points in thedirection of the flow. This installa-

tion is calculated as if the ther-mowell is subjected to flow alongits entire insertion length (L). Incase 2, the thermowell is subjectedto a flow from its tip; a positionwhich is preferable for many appli-cations. Here, the calculation isalso carried out over the entire in-sertion length, however, the calcu-lation of the static bending mo-ment is outside of the scope ofASME PTC 19.3 TW-2016. In thiscase, the determination of thebending moment and the Strouhalnumber should be made on the ba-

1.

2.

Diagram 2: Optimis-ation of the fre-quency ratio, per-centage illustration

• Meet the Wikaexperts at this year’sHannover Messe(April 24–28, 2017):Hall 11, Stand C56.

PROCESS-Tip

Sources:Wika

Diagram 1: VariableStrouhal number ofthe PTC 19.3 TW-2016

L e n g t h - r e l a t e d i l l u s t r a t i o n o f t h en um b e r o f c y c l e sINSERTION LENGTH L NATURAL FREQUENCY FN DURATION FOR 1011 CYCLES

250 mm approx. 300 Hz approx. 5 years

350 mm approx. 150 Hz approx. 10 years

450 mm approx. 100 Hz approx. 17 years

550 mm approx. 60 Hz approx. 26 years

document2483191793277848697.indd 27 27.02.2017 14:54:35

PROCESS AUTOMATION

28 PROCESS Worldwide 1-2017

sis of computational fluid dynam-ics (CFD) or experimental measure-ments.For an angled installation of a

thermowell into the pipeline (seeFig. 2), the standard has been ex-panded by a chapter (6-10.8). Withthe calculation, the direction offlow does not matter—the ther-mowell will be designed in accord-ance with a conservative method,as though it was installed perpen-dicular to the process flow.

Requirements for theminimum tip thicknessof the stem

The tip thickness of the stem (t)within the dimensional limits inaccordance with ASME PTC 19.3TW-2016 has now been defined asat least 3.0 mm (chapter 4.1).This change has enormous rele-vance, primarily for the optimisa-tion of the thermowell in theevent of a failed calculation. Therequirements contained in “TW-2010” state the tip thickness mustat least correspond to the mini-mum wall thickness of the ther-mowell at the tip. This led to largeproblems in the everyday applica-tion.To understand this, one must

bear in mind which possibilities foroptimising the thermowell designare actually PTC 19.3 conformant(see Fig. 3). Plant operators caneither shorten the insertion length(L) or strengthen the root diameter(A) and the tip diameter (B) of thestem that is subjected to the flow.However, for these diameters andthe taper ratio B/A resulting fromit, the specifications of table 4-1-1in chapter 4.1 set limits.For example, the taper ration

B/A must be within 0.58 (= max.taper) and 1 (= straight stem). Inpractise however, these regula-tions are not without problems:Thus a strengthening of the rootdiameter is limited by the innerdiameter of the flange nozzle. Anincrease in the tip diameter, on theother hand, has a negative effecton the response time of the ther-mometer.Diagram 2 illustrates the effec-

tiveness of a change of root and tipdiameters with the example of astraight thermowell with a stem of16 mm diameter:

First the root diameter is in-creased to 27 mm (= max. taper).In the example case, this measurewould improve the ratio of theStrouhal frequency to the naturalfrequency (fs/fn) by 51 percent,without considerably diminishingthe response time.An increase in the tip diameter,

carried out in a second step, to27 mm (= straight design) wouldyield an additional optimisation ofthe frequency ratio by 13 percent,however at the price of a clearlylonger response time.Using the previous standard, its

value would deteriorate further,since the tip thickness would ac-cordingly have to be increased to10 mm.In addition, the sensor length

would need to be adapted to matchthe changed bore depth. In accord-ance with the current ASME PTC19.3 TW-2016, however, the tipthickness could remain unchangedso long as its thickness is at least3 mm.

Further changes

The revision of the TW sectionof PTC 19.3 has also been used byASME to simplify the understand-ing and readability of the standardthrough a modern style of illustra-tion as well as to clean up therounding errors in the calculationexamples.

Summary

The ASME PTC 19.3 TW-2016standard offers users improve-ments in key details. But as withthe writing of computer programs,also with wake frequency calcula-tions the motto “garbage in, gar-bage out” always applies. Thismeans that the results of thestrength calculation can only be asgood as the process data on whichthey are based as input parame-ters.If, for example, the flow rate of

the medium in the process variesby ±20 percent, the relevant re-sults, such as the frequency ratio,will show the same level of varia-tion. This variation, in turn, makesa serious evaluation of how thethermowell must be designed im-possible.

Fig. 1: Differentflows in pipe el-bows

Fiig. 2: Angledinstallation in apipeline

Fig. 3: Tip thickness“t” in accordance toFig. 4-1-1 of thePTC 19.3 TW-2016

Sources:Wika

3.

document2483191793277848697.indd 28 27.02.2017 14:54:38