Prestress Loss in Pre-tensioned beams

[subho.nit]

I am modelling a 3 span bridge with Precast pretensioned beams with integral connection at Pier location & bearing support over the end abutments. The prestressing system has been defined as
let#AZ 150 $ Cable Area (mm2)
let#ZV 0.751860#AZ/10^3 $ Prestress force per tendon (kN)

let#MU 0.0 $ fricition coeff.
let#BETA 0 $ wobble factor [°/m]
let#KS 0 $ slip[mm]
let#ECC 0 $ excentricity of tendon within duct [mm]
let#MINR 1 $ minimum bending radius [m]
let#DO 15.7 $ external diameter [mm]

$ PREESTRESSING DEFINITION
SYSP NOPS 1 NO 0 MAT 4 ZV #ZV AZ #AZ LITZ 1 MINR #MINR BETA #BETA BETG #BETA MUE #MU MUE- #MU $$
ECC #ECC CDIS 40 CEDG 150 SP #KS SC #KS SF #KS VP 0 VC 0 VF 0 LPRE 0 LCOP 0 LFIX 0 $$
DO #DO DI #DO TITL “Pretension T15”

I’m not sure whether the system is considering the Elastic deformation loss in the system. I’ve tried with different combination of "CS " definition in the tendon module, i.e., CS ICS1 2 ICS2 3 (prestressing activation- CS2, grouting - CS3), CS ICS1 2 ICS2 0 (for immediate bond), CS ICS1 2 ICS2 2 (same stage grouting/bonding) but in any of the case expected loss (~5-6%) couldn’t be achieved. Also, the segregation of losses (Elastic deformation, Short term relaxation/Long term relaxation, Short/Long term creep/shrinkage) I’m not able to extract. The overall loss I’m getting is around 12-13% which is quite low in comparison to the general loss expected in PSC members. I am designing as per Eurocode 2 & for creep/shrinkage and relaxation losses, considering the default values.

Can anyone suggest on these?

[JFH]

Hello

The different parts of the stress losses can be obtained from Graphic.
I have already described the procedure in another post.

Creep and shrinkage for quad elements

Hello The effect of creep and shrinkage can be obtained from a QUAD-system. You have to look at the 4XXX or 5XXX load cases. I will do the calculation using the following TEDDY example. TEDDY > File > Examples > csm > english > csm4_quad_singlespan.dat Open the Graphic (Wingraf) and visualise the tendon stresses (or forces) [grafik] The difference between the 4XXX load cases describe the losses due to relaxation, creep and shrinkage. For example: (1) load case 4011: 1279,73 MPa (on the …

Apart from that, the workflow to define the construction phases is explained in our online bridge design tutorial.

Bridge Design Training #10 - Construction Stage Manager (CSM)

You will also find some useful examples in the TEDDY.

TEDDY > File > Examples > csm > english > more > e.g. csm3_casting_bed_method.dat.

Best regards
Frederik Höller
Your SOFiSTiK Support Team

[mico]

Hi,

With ICS2 0 always master cross section is used, so tendons will also get compresive force due to prestress which is loss due to elastic shortining indeed.

In pretensioned beam elastic shorting losses are main losses. Try doing LC with PT before applying deadload (f.e. with ICDS pt_cs+1) and in pt_cs you see stresses in tendons in stage when only load is prestress, so it should be intial value recuded by el.sh.

[subho.nit]

Hi Mico/Hi Frederik,
Thanks for your suggestions.
@mico , I have tried with ICS2 0 but it doesn’t include any elastic deformation loss. Pasting some snips from the model

this is the force in beam at the stage of prestressing. The initial force given : 16 No.s 150 mm2 tendon. Total : 3348 kN. Sofistik calculated force 3264 kN. which is 2.5% less than the applied one. The sofistik result output shows :

I’ve considered the Prestress stage before the dead load application. CSM input is given below
+PROG CSM urs:65.1
HEAD Construction stage analysis
CTRL ASE TEXT ‘CTRL WARN 827’
CTRL CANT 11
echo opt full val yes
$CTRL GAMC 1.35
CTRL SVRF 0
CTRL EIGE EN10
CTRL Aqb text ‘echo eige extr’

CS 1 TYPE “G_1” TITL “Deck Iso Cast In Yard - Element Activation”

GRP 50,51,52,60,61,62,70,71,72 ICS1 1 ICSD 1 T0 3 PHIF 0 $ Springs
GRP 105 ICS1 1 ICSD 1 T0 3 PHIF 0 $ Rigid Link
GRP 1,2,3,4 ICS1 1 ICSD 4 T0 10 PHIF 0 $ Girder

CS 2 TYPE “P” TITL “Prestress”

CS 4 TYPE “G_1” TITL “Beams Weight”

CS 15 TYPE “C_1” TITL “CRSH” T 20 RH 80 TEMP 20 NCRE 2

CS 30 TYPE “C_2” TITL “CRSH” T 60 RH 80 TEMP 20 NCRE 5
CS 40 TYPE “C_2” TITL “CRSH” T 120*365 RH 80 TEMP 20 NCRE 5

END

Please suggest if I’m missing something.

@JFH I’ve raised for this issue multiple times with sofistik user support (through mail) but didn’t get any clear answer for the elastic deformation particularly. Can you please illustrate how to take elastic deformation loss in the pretensioned beams? Also, 7002 load case shows the beam forces as 3264 kN

Final stage load as 2869 kN , Loss : 395 kN.

Sum C results shows loss of 395 kN for the beam. so no losses for long term relaxation?

Can you please clarify?

[mico]

Hi,
what I suggest You to check is stress in tendons in LC 7002 (min/max) in wingraf/design/stress…
I guess total normal force in beam may be represent average losses if you have small amount of wire in top of the section and high in the bottom thier losses are going to differ casing bending moment.

[subho.nit]

Hi,
@mico I have checked the load cases for Tendon stress below are the snip :

This is the minimum stress in tendon for the Applied prestress loadcase (7002).

This is the maximum stress in tendon for the Applied prestress loadcase (7002).

The initial applied stress is 0.75x1860 = 1395 MPa. Can we get average prestress in the tendons? Considering the maximum stress, we are getting (1395-1364)/1395x100 = 2.22% loss only. Am I interpreting correct?
But in any case, couldn’t get the expected result.
If the elastic deformation losses are there, then it should be reflected in the beam axial forces, isn’t it? but it is not getting reflected as I had pasted snips in my earlier post.

[mico]

Losses differ between tendons in a section, because they are proportional to concrete compressive stress around tendon. If stress distribution on your section is like 0 MPa in the top and 25 MPa in the bottom, you are going to get no losses in top tendons and around 10% in bottom ones. Normal force is sum of tendon forces after losses.

1271/1395=91% so you have max loss ~9%.

Try to modify Ec of concrete or Es of steel to manipulate losses, you will see differences in result with diffrent modulus if elastic deformation is accounted.

One more thing which I am not sure, but I think sofistik doesn’t account automatically age-adjusted Young modulus in ASE.

[subho.nit]

Hi @mico , I’ve seen that if we consider the minimum stress in the tendon, it is ~9%. But we can not get the average stress and that’s why I was more inclined towards the beam axial forces where we can get the overall picture.
I am rather worried regarding the possibility that you have mentioned : “but I think sofistik doesn’t account automatically age-adjusted Young modulus in ASE.” @JFH Can you please confirm whether it is so? If it is so, there is no point in mentioning the “T0” in CS definition in CSM. Also, can you please share an example where this Elastic deformation has been considered for pre-tensioned beam? I’ve referred the earlier example mentioned by you “csm3_casting_bed_method.dat.” but didn’t quite find out this particular issue.

[mico]

T0 in csm is provided for creep and shrinkage calculation.
Average force is normal force in beam.

To check if sofistik account T0 for young modulus you can try 2 options with t0=3 and t0=28. From my experience there is no difference in results.