Handling and Shipping of Long Span Bridge Beams George Laszlo, P.E. Chief Engineer Morse Bros. Prestress Concrete Group Clackamas, Oregon Richard R. Imper, P.E. Vice President; General Manager Morse Bros. Prestress Concrete Group Clackamas, Oregon n the early history of the prestressed concrete industry (up to 1960), bridge beams were very bulky (heavy) and their length seldom exceeded 100 ft (30.5 m). In fact, at that time, 80 ft (24.4 m) beams were considered quite long. With advancing prestressing technol- ogy, together with the improvement of materials (especially high strength con- crete and high strength prestressing steels), beam sections became progres- sively more structurally efficient. As a result, beam sections became more slender and the spans much longer. Currently, on the West Coast, 148 ft (45.1 m) beams are common and 130 to 140 ft (39.7 to 42.7 m) bulb T beams and I beams are used frequently. Today, the length limit of bridge members is de- termined mainly by the mode of trans- portation (truck steering trailer) and al- lowable gross weight rather than any ar- bitrary span restriction. Over the years, as beam sections be- came more slender and their spans longer, producers soon discovered that these long beams had a tendency to crack or even collapse during handling or shipping unless the lifting points (shipping points) were moved away from the ends of the members, or special braces were attached to the beam. The lateral stability of these types of beams was discussed in the 1960s (see Refs. 1, 2 and 3). Further information on this topic may also he found in Refs. 4 and 5 (published in 1971). Currently, the PCI Design Handbook, in Section 5.2.9, Lateral Stability, briefly describes the problem and suggests so- lutions. Ref. 4 is the basis for this sec- tion. Lateral instability occurs during the handling and shipping of long pre- stressed bridge beams. This problem arises because of the imperfections 86
16
Embed
Handling and Shipping of Long Span Bridge Beams - pci.org · PDF fileHandling and Shipping of Long Span Bridge Beams George Laszlo, ... n the early history of the prestressed ... section
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Richard R. Imper, P.E.Vice President; General ManagerMorse Bros. Prestress Concrete GroupClackamas, Oregon
n the early history of the prestressedconcrete industry (up to 1960), bridge
beams were very bulky (heavy) andtheir length seldom exceeded 100 ft(30.5 m). In fact, at that time, 80 ft (24.4 m)beams were considered quite long.
With advancing prestressing technol-ogy, together with the improvement ofmaterials (especially high strength con-crete and high strength prestressingsteels), beam sections became progres-sively more structurally efficient. As aresult, beam sections became moreslender and the spans much longer.
Currently, on the West Coast, 148 ft(45.1 m) beams are common and 130 to140 ft (39.7 to 42.7 m) bulb T beams andI beams are used frequently. Today, thelength limit of bridge members is de-termined mainly by the mode of trans-portation (truck steering trailer) and al-lowable gross weight rather than any ar-bitrary span restriction.
Over the years, as beam sections be-came more slender and their spanslonger, producers soon discovered thatthese long beams had a tendency tocrack or even collapse during handlingor shipping unless the lifting points(shipping points) were moved awayfrom the ends of the members, or specialbraces were attached to the beam.
The lateral stability of these types ofbeams was discussed in the 1960s (seeRefs. 1, 2 and 3). Further information onthis topic may also he found in Refs. 4and 5 (published in 1971).
Currently, the PCI Design Handbook,in Section 5.2.9, Lateral Stability, brieflydescribes the problem and suggests so-lutions. Ref. 4 is the basis for this sec-tion.
Lateral instability occurs during thehandling and shipping of long pre-stressed bridge beams. This problemarises because of the imperfections
86
during production (slight horizontal ec-centricity in prestress from the Y-Y axis,or the lifting loops not centered in thesection or thermal gradient from oneside of the beam to the other). Such vari-ations cause the beam to bow horizon-tally during handling, thus shifting thecenter of the mass away from the origi-nal centroid of the beam. When thebeam is lifted, the combination of thehorizontal bow with the tilting actioncauses the beam to bend and deflectprogressively in the weak axis.
As soon as the moment of inertia of theweak axis becomes cracked, thephenomenon rapidly increases until thebeam fails in compression or tension,depending on the configuration of thebeam in the Y-Y axis. Such a phenome-non may occur even where no lateralloads or direct forces are applied.
From the above brief discussion andRefs. 4, 5 and 6, it is shown that failuredue to lateral instability is related to de-flection:
4
A = K u'— (1)E,I„
where K = 1/120 for the mass center of asimply supported prismatic beam.
Assuming that the design engineerhas already established the beam crosssection and prestress level, the optionsavailable to the producer to reduce thedeflection A (and thus improve lateralstability) are as follows:
1. Shorten the handling and shippingspan.
2. Improve the modulus of elasticityof concrete (Er) (increase concretestrengths f, { and f, ).
3. Brace the member (thereby in-creasing the effective I i,).
The above options are now brieflydiscussed:
1. Shortening of the handling andshipping span can be impractical be-cause most engineers design bridgebeams for in place load and in placesupports. Since the economical design
Synopsis
The stability and handling stressesof long span prestressed concretebridge beams are discussed. Specialattention is given to the location of
lifting points, shipping supports, effectof road conditions (impact) andsuperelevation (beam tilt). Variousbracing systems are discussed andnew proposals are put forth.
A suggested analytical procedure is
presented whereby a long span bridge
beam can be designed for stabilityduring handling and shipping. To il-
lustrate the method, a sample designof a 136 ft (41.5 rn) long PCt 72 in.(1829 mm) bulb T bridge beam forhandling and shipping is included.
for a simply supported beam requiresmaximum allowed compression in thebottom fiber and maximum allowed ten-sion in the top fiber at the midspan ofthe beam due to its own dead load attransfer stresses, it is extremely difficultin some cases to shift the lifting pointsaway from the end without significantlyincreasing the concrete strength.
2. There is a practical (economical)limit to increasing the concrete strengthand improving E, especially for earlytransfer strength and for handling of thebeam from the casting bed.
3. Bracing the member has some meritand was the most commonly usedmethod in handling and shipping beamsin the past.
Each producer developed a bracingsystem to fit the particular plant re-quirements. The basic method includedstiffening the weak part of the beam(most often the top flange) with pipeframe, temporary trusses or othermethods. Since every long beam has tobe stiffened, this type of temporarybracing system can be very expensive.
PCI JOURNAL1November- December 1987 87
BEAM PLAN x/CABLE STAY
STRANDANCHOR
41'R HOLE KINGPOST
SECTION AT END BLOCK SECTION AT MID-SPAN
Fig. 1. Bracing system (king post truss) used for shipping long span bridge beams.
For shipping the beams the mostcommonly used bracing system is theking post truss system (see Fig. 1),where the posts are attached to the beamsides at midspan and several prestress-ing strands are stressed over the kingpost and anchored near or at the end ofthe beam. Some producers have usedheavy cable up to 1 in. (25.4 mm) diam-eter instead of prestressing strand, theadvantage being that the cables can beadjusted by heavy duty turnbuckles.
During shipping the beams may besubjected to their most severe stressconditions. Impact forces depending onthe road conditions can be in the rangeof 20 to 30 percent in both directions (upor down). The superelevation in theroad curves or job site conditions may beas much as 10 percent. Therefore, thehorizontal component of the beamweight and the weak axis bending mo-ment are increased accordingly. Thecombined stresses due to the prestressand horizontal bending may reach wellbeyond the cracking or compression
limit of the concrete (see Fig. 2).Unfortunately, the king post and
strand system has limitations. Morethan one prestressed concrete beamproducer learned the hard way, that be-yond a certain horizontal deflection ofthe beam, the system is useless and can-not prevent the collapse of the beam.
Most producers used the king poststrand truss system with one or twostrands on each side of the beam. Theanalysis of the statically indeterminatecomposite king post truss, by themethod of consistent deformation,would indicate that for cases of longbeams more than one or two strands (orcables) would be needed to preventcritical lateral deflection.
On the other hand, this system couldbe effective for short durations (movingon the superelevated road where thecurve changes) because the beam de-flection is not instantaneous. It isgradual and can be observed visuallydue to the plastic characteristic andcreep of the concrete.
The situation becomes critical whenthe truck must stop and stay on thesuperelevation (or on an uneven sur-face), where the gradual horizontal de-flection can not be adjusted or stopped.
One of the authors observed a 130.5 ft(40 m) long, 84 in. (2134 mm) deep Ibeam which was lifted from the trucktrailer for erection and later had to bereplaced because the crane could nothandle the load. Upon being replaced,due to road superelevation and thebeam offset on the truck, the beam hadabout 15 percent (8.87 degrees) tilt. Ittook about 20 minutes to reach the criti-cal deflection, about 12 in. (305 mm)
horizontal bow, before the beam col-lapsed. (This beam had five strands oneach side of the king posts.) Other pro-ducers across the country have hadsimilar experiences.
Our company re-evaluated the beamhandling and shipping problem duringthe past two years. The final conclusionwas a simple one.
Handling the beam in the plant mustbe done under very strict conditions andconstant supervision. Impact and tiltingforces can be avoided whereby theanalysis given in the PCI Design Hand-book to evaluate stability can beapplied directly.
PCI JOURNAL/November-December 1987 89
YUKE
BEAM E
iz a
N ^
O y:
N.A
Fig. 3. Beam lifting device used to increase top fiber distance (y = ) of beam.
For simply supported beams, use Eqs.k5.2.2) and (5.2.3):
5 wt(2)
^L ° 384 E,1 11
with a factor of safety of:
F.S. = y` _- 2 (3)13V
Note that Eq. (5.2.2) includes an ad-ditional safety factor because the truemass shift is:
w1^(4)/3M= 120
Therefore, the true safety factor andEq. (5.2.3) become:
F.S. =
(5)9.64(^
For the details, see Ref. 5.Since the factor of safety is a function
of y, and /3,,, it is logical to increase y, orreduce f3.
It is possible to increase the vertical
distance between the center of mass andIiffting points (y t ) by using a rigid yoke.This technique is sometimes used inplants (see Fig. 3).
As mentioned earlier, the modulus ofelasticity of concrete (Er), i.e., the con-crete strength, may be increased or thelifting points shifted away from theends. In this case the stability factor atmidspan may be expressed as:
— (51 z – 24a^) (6)
" – 384 E^ I v
Fig. 4 shows the relation beween theall ratio and deflection.
Based on plant and field experience(and the actual value of j3„), we use thefollowing safety factors:
1. For plant handling: F.S. > 1.52. For field handling (erection): F.S. >
1.75For transportation stability, the sup-
port (bunking) points must be knownand the impact factors estimated (20 to30 percent recommended). It is impor-tant that the roadway superelevations(especially at the jobsite) be checked
90
w C^ 5 w V
M g 384 E
le
w
a $ = zc
MB (€- 4a)
J- 15e 1 244 )384E1
IDo
90
Ba
70
a60U.050
040
X30
0z
20
10
Wa 0
EXAMPLE:100 Ft. END SUPPORTED BEAM Q-I.00 in
SAME BM. SUPP. 6 Ft FROM ENDS ( .5B in.
0.02 0.04 0.06 0.08 0.10
a/L
Fig. 4. Design aid for determining beam support influence on deflection.
and that the beam lie on a level surfaceand vertical position prior to erection.
Using the above assumptions, thecombined stresses of a laterally de-flected beam at any point can be calcu-Iated.
It can be quickly determined that thecritical points are the downward topflange under high tension and the up-ward bottom flange under high com-pression at the midspan of a bulb T or Ibeam (see Fig. 2). These stresses can becounteracted or limited by mild steelreinforcement or prestressing steel, ifthe stresses are below cracking, the sec-tion may be conventionally reinforcedwith supplementary reinforcing barsnear the outer edges of the flange. If thestresses are above the allowable, tempo-rary post-tensioning may be used ad-vantageously to reduce them to an ac-ceptable level (see Fig. 5).
Our experience in the past 2 yearsshows that, by using nearly the same lo-cation for shipping as for yard handling,highly stressed 72 in. (1829 mm) bulb Tbeams in the 140 to 150 ft (42.7 to 45.7m) range may be safely shipped with
Fig. 5. To reduce high tensile stresses,temporary post-tensioning may be used tobrace a long slender beam.
PCI JOURNALMovember-December 1987 91
Table 1. State of Washington DOT specification for stability.
CRITERIA FOR CHECKING GIRDER STRESSES AT TIME OFLIFTING OR TRANSPORTING
Stresses at both support and harping points must be satisfied based on:1. Specify concrete strength at time of Idling or transporting, f,,^
compressive strength at time of lifting or transporting verified by testbut shall not exceed design compressive strength (f,) at 28 da ys inpsi +1000 psi.
a. With no bonded reinforcement = 3 , fb. With bonded reinforcement to resist total tension force in the concrete
computed on the basis of an uncracked section = 7.5 f ,'mThe allowable tensile stress in reinforcment is 30 ksi (ASTM A 615Grade 60).
4. Prestress losses1 day — 1 month = 20,000 psi1 month — 1 year = 25,000 psi1 year or more = 35,000 psi (max.)
5. Impact10 to 30% depending on road and field conditions, acting up or down.
Fig. 6a. Handling of long span beam at plant.
92
0
r
4 1
Fig. 6b. Long span beam being braced prior to shipping.
t
Fig. 6c. Transportation of long span beam.
PCI JOURNALNovember-December 1987 93
Fig. 7. Collapse of long span beam.
only two, or a maximum of four, un-bonded post-tensioned strands placed inthe top flange. These strands can be re-leased and recovered after erection ofthe beams.
We use '/z in. (13 mm) diameter 270Klow relaxation strands in a greased plas-tic tube, anchored with steel plates andstandard barrel anchors (chucks) at eachend of the top flange. The stress is re-leased by a hydraulic jack, or by burningthe strands through a small hole pro-vided in the flange about 6 ft (1.83 m)from one end, After releasing the strand,it is pulled out manually.
For shorter beams [less than 130 ft(39.6) ], only additional mild reinforcingsteel may be required in the top flange.
Both methods are simpler to use, easyto calculate, and much less expensivethan any other technique we have usedin the past.
Only a few codes of practice addressallowable stresses and other parametersin handling and shipping of long pre-stressed concrete beams. For example,
AASHTO, under "Allowed TemporaryStresses," specifies 0.6 f,' for compres-sion and 7.5 r f,' for tension.
Table 1 shows the State of WashingtonDOT specification on criteria for check-ing girder stresses at time of handling orshipping.
Figs. 6a through 6c show the handlingand shipping of a long span beam.
Fig. 7 shows a vivid example of thedisastrous collapse of another long spanbeam. The importance of followingsound handling and shipping proce-dures for long span members cannot heover-emphasized.
Appendix A presents a step by stepanalytical procedure for dealing withstability when handling and shippinglong span prestressed concrete bridgebeams.
Appendix B gives, with the aid of anexample, the detailed calculationsneeded for carrying out such an analysis.
Appendix C summarizes the meaningof each mathematical symbol used inalphabetical order.
94
CLOSING REMARKSThe proposed procedure presented in
this article is a straightforward methodfor dealing with three troublesome con-ditions involved in handling and ship-ping long span prestressed concretebridge beams:
(a) Safe handling at the time of strip-ping when concrete strengths areat their lowest and forces in theprestressing strand are at theirhighest.
(h) Safe truck delivery with the un-certainties of superelevated roadsand dynamics of travel.
(c) Adequate safety factor duringerection where handling of pre-cast members by a contractor maynot be as well controlled as theprecaster might assume.
Conditions (a) and (c) represent theclassical lateral stability problem whenthe beam is hanging from pickup loopsfrom the top of the beam. The methodgiven in the PCI Design Handbook,Section 5.2.9, is used with new pro-posed safety factors together with a de-sign aid shown in Fig. 4. Lateral stabil-ity is dramatically improved by movingthe pickup points in from the ends of thebeam, subject to avoiding overstress inthe beam. This is particularly critical inthe net top flange tensile stresses at theharp point near 0.41.
Condition (b) represents a combinedstress analysis of vertical bending of the
beam on truck hunks, plus lateralbending due to an assumed superele-vated road. In critical cases, surveyingthe road and measuring superelevationis important. When combined stressesexceed allowable values for short termloading, the use of temporary post-ten-sioned strands in the top flange, andlorhigher strength concrete are shown tobe effective in improving the safety ofthe beam.
REFERENCES1. Muller, Jean, "Lateral Stability of Precast
Members During Handling and Placing,"PCI JOURNAL, V. 7, No. 1, January-February 1962, pp. 20-31.
2. Swann, tt. A., and Godden, W. G., "t'heLateral Buckiing of Concrete BeamsLifted by Cables," The Structural En-gineer (London), V. 44, No. 1, January1966, pp. 21-33.
3. Laszlo, George, "A Prestressed Light-weight Concrete Bridge 131 ft. Long,"Civil Engineering - ASCE, V. 37, No. 4,April 1967, pp. 64-65,
4. Anderson, Arthur R., "Lateral Stability ofLong Prestressed Concrete Beams," PCIJOURNAL, V. 16, No, 3, May-June 1971,pp. 7-9.
5. Swann, R. A., Readers' Comment to "Lat-eral Stability of Long Prestressed Con-crete Beams" (Ref. 4), PCI JOURNAL, V.16, No. 6, November-December 1971, pp.85-87.
6. PCI Design Handbook, Third Edition,Prestressed Concrete Institute, 175 W.Jackson Blvd., Chicago, Illinois, 1985.
* w ,r
METRIC (SI) CONVERSIONFACTORS
1 ft = 0.305 ni 1 psi = 0.006895 MPa1 in. = 25,4 min 1 kip-ft = 1356 N-m1 pcf = 16.02 kg/m3
1 kip = 4448 NI lb/ft = 1.488 kg/m
PCI JOURNAI /November-December 1987 95
APPENDIX A - CALCULATION STEPSDetailed below are the calculation steps needed to carry out a stability check
for handling and shipping long span bridge beams.
Plant and Field HandlingI. Select the factor of safety (F.S.).
Note that minimum F.S. for planthandling = 1.5; for field handling =1.75.
2. Calculate the beam deflection as asimply supported span in the Y axis;
5wl°384 E,. I„
3. Compute y,/F.S. = # in order to findthe optimum beam stability factor(deflection).
4. From the ratio of /3„f , find all fromFig. 4, which gives the optimumcantilever and center span lengths.
5. Round out the center span and can-tilever lengths to the nearest foot.Check the safety factor with theconventional deflection formula atmidspan for equal cantileveredbeams in the Y axis:
Wi aw= 384 ELI„ (51' – 24x2}
With the new support locations,check the handling stresses at criti-cal points (at the harp point which isgenerally near 0.41 and support).Check the required concretestrength and revise the value ac-cordingly,
Beam Shipping (Handling)8. Establish the transportation support
points. It is prudent to check thisoperation prior to yard handling andshipping.
9. Check the stresses at critical loca-tions (at the harp point which isgenerally near 0.41 and support)with (a) no impact and (b) impact upand down.
10. Check the road superelevation orassume an arbitrary beam tilt. Cal-culate the weak (Y-Y axis) momentdue to beam tilt,
11. Calculate the top and bottom flangestresses due to beam tilt only.
12. Combine (add) the shippingstresses (Step 9a) to the flangestresses (Step 11). Note that no im-pact is assumed at combinedstresses. It would be extremely con-servative to add the most adverseconditions unless the road is ex-tremely rough at superelevations.
13. If the combined stresses at anypoint exceed the allowable stresses,post-tension or reinforce the re-quired part of the section.
14. If stresses are below cracking, re-inforce the section according tousual stress relations (compres-sion-tension).
15. If combined stresses are above thecrack level, use post-tensioning.
16. When supplementary strands areadded, revise (a) total prestress forceand (b) prestress eccentricity.
17. With the new prestressing force (P)and eccentricity (e .), check thestresses at the critical locationssimilar to Step 9.
18. Combine these stresses (Step 17)with the stresses in Step 11.
19. Check the concrete strength, in-crease it if required, or add morepost-tensioning.
20. Post-tensioning can be very helpfulas the need arises. For example, if avery high stability is required, orthe concrete strength cannot he in-creased, post-tensioning can beapplied before yard handling (inthe prestressing bed) and can thenbe postponed until the beam iserected.
96
APPENDIX B - SAMPLE CALCULATIONS
SECTION PROPERTIESRELATIVE T0: X AXIS Y AXIS
PERIMETER 254.8 frAREA 767 In2CENTEROID 21 inYb 36,60 in 42 TOP FLANGE WIDTHYt 35.40 in 26° STM, FLANGE WIDTHlb 14915.0 In3 1792.101 Zy TOP FLANGEZ t 15421.0 fns 2894.9 Zy BTM, FLANGEI 545894 Vn4 37634.13 [ y
fCi= 4500 psi L. TRANSFERfc= 5500 psi Q 28 DAY
I3-0 C.G.S. 44- Z "♦ 270 K LO
W
LAX" STRANDS
ti ao D
54'..4"
PCI BT-72 BRIDGE BEAM (7 FT. SPACING 8" DECK)
Given: PCI BT-72 bridge beam withfollowing data (see also diagram):Unit weight of concrete = 155 pefw = 825.6 lb/ft1 = 136.0 fta=oftEe=4.07x106psif= 4500 psiI= 37634 in.4
X =136,
A. BEAM HANDLING INPRODUCTION YARD
Step I — Select factor of safety(a) Assume factor of safety (F.S.) = 1.5
Step 2 — Calculate beam deflection.
5 w14_
384 E,I„5 x 825.6 x 136 4 x 1728
384 x 4.07 x 10 6 x 37634= 41.48 in.
Step 3 — Compute stability factor.
Ru =yr_ 35.4 = 23.6 in.1.5 1.5
Y 23.6=0.57
41.48
Step 4 — Determine optimum can-tilever and span lengths.
For,Q„/A„= 0.57, all = 0.063 (see Fig. 4).Therefore,a= 0.063x136=8.57ft1,= 136-2x8.57 = 118.868
(10 percent losses)P= 44 x 28 = 1232 kipsex^^F r„o = d–yr= 67 - 35.4 = 31.6 in.
frou= + 1232 _ 1232 x 31.6
0.767 15.421+ 12 x 1326.4
15.421= + 114 psi (compression)
1232 1232 x 31.6fboa^"` _ + +
0.767 14.9I5_ 12 x 1326.4
14.915= + 3149 psi (compression)
Step 7 — Check required concretestrength.
Required handling strength:3149
= 5250 psi0.6
New factor of safety for modulus of elas-ticity of concrete, 4 (f j = 5250 psi)4.39 x 10 psi
('
New F.S. = 1.55 x 4 '39 = 1.674.07
Note: If cantilever is rounded to 8 ft,new F.S. would be 1.53.Stresses at pickup point and midspanmay be calculated in a similar way (usu-ally will not govern).Conclusion: If beam is handled 9 ft fromends, then concrete strength must be5250 psi > 4500 psi release.
B. BEAM SHIPPING (HAULING)
Step S - Determine transportation sup-port points.Assume the same support points as forhandling, i.e., 9 ft from each end.f,' = 5500 psi; E, = 4.496 x 10 6 psiP = 27.17 kips/strand (25 ksi losses)P = 44 x 27.17 = 1195.48 kipsImpact = 20 percent up or downSuperelevation = 8 percent
Step 9 - Evaluate stresses at critical lo-cations.
(a)Check stresses at 0.4 1 (harping point)without impact,
= 1326.4 kip-ft (see Step 6)_ 1195.5_ 1195-5 x31.6
(b) Check stresses with 20 percent im-pact up.Ar um = 0.8 x 1326.4
= 1061 kip-ftf op = + 65 psifrlo„ = + 3238 psi
Required f,' = 3238 - 5397 psi
0.6< 5500 psi (ok)
(c) Check stresses with 20 percent im-pact down.Values will he less than the abovestresses, due to increased beam weight.Check at support:d= [r 67-42 lx9 ] +42