Wind Foam Cutters Bades Ect

8/13/2019 Wind Foam Cutters Bades Ect

1/24

SAND REPORT

A

SAND2002-2424Unlimited ReleasePrinted August 2002

Evaluation of Design Concepts forAdapt ive Wind TurbineBlades

:on tirtntrial Energ ..--pts, LCKirkland, Wa gton 98109Prepared by PSandinNationalLaboratoriesAlbuquerque. New Ma- 87185 and Livermore,California 94550Sandia is a rnuitiprcgram laboratory operated by andi Corporation.hkheed tdartin Campany. orthe United States hpar tman t ofnder Con t r a c t DE-AC04-WAL85OM); ~Approved tor public release; firmer dissemination unlimited.


2/24


3/24

SAND2002-2424Unlimi ted ReleasePrinted August 2002

Evaluat ion o f Design Conc epts fo rAd apt ive Wind Turb ine B lades

Global Energy Concepts, LLC5729 Lakeview Drive NE, IO0Kirkland, Washington 98109

Principal Investigator:Dayton A. Grif fin

AbstractThe objectives of his work were to develop conceptual structural designs for an adaptive (bend-twistcoupled)blade,o valuate andidatedesign oncepts,odentify onstraints nd/or oncernsormanufacturing, load paths, and stress concentrations, to develop estimates of structural performance andcosts, and to select a single configuration as showing the greatest potential for success in manufacturing,strength and durability.This report summarizes the work performed on this project, including the approach taken, configurationsand materials considered,and the computational methodology used. The work presented includes:

Candidate fiber orientations and fabric architectures for adaptive blade manufacture are identifiedand assessed on the asis of estimated static strength, stiffness, and fabrication costs.A parametric study is performed for potential blade structural arrangements. Each configuration isevaluated on the basis of estimated manufacturing cost and magnitude of bend-twist couplingachieved.Based on the parametric study results, a single configuration is selected for hrther evaluation. Acomplete blade model is developed and assessed on the basis of estimated manufacturing cost andbend-twist behavior under steady loading.


4/24

AcknowledgementsSandia TechnicalMonitors:Thomas D. AshwillPaul S. Veers

This is a Contractor Report fo r Sandia National Laboratories that partially fulfillsthe deliverables under Contract #21010.


5/24

Table of Contents1. Technical Approach ....................................................................................................................... 6

1.1Overview .................................................................................................................................. 61.2 Approach ................................................................................................................................. 61.3Concepts nvestigated ............................................................................................................... 7

1.3.2 Conventional Triaxial Skins with Biased Box Spar Caps ................................................... 71.3.3 Biased Skins and Biased Spar Caps ................................................................................... 7

2. Candida te Com posite Materials..................................................................................................... 82.1 Stitched HybridFabric .............................................................................................................. 82.2 Oriented Non-Continuous Carbon Fibers .................................................................................. 9

1.3.1 Biased Skin with Conventional Unbiased Spar Caps .......................................................... 7

3. Parametric Studyof Candidate Structural Configurations.......................................................... 103.1BaselineSectionStructuralDefinition ..................................................................................... 103.2MaterialProperties ................................................................................................................. 113.3 Calculation of BladeSectionStructuralProperties .................................................................. 123.4 ParametricStudyResultsandDiscussion ................................................................................ 13

4. Complete Adaptive Blade M odel ................................................................................................. 155. Conclusions................................................................................................................................... 18 6. References .................................................................................................................................... 19


6/24

1. Technical Approach1 I verviewThe objectives of this work were to develop conceptual structural designs for an adaptive (bend-twistcoupled)blade,o valuate andidatedesign oncepts,odentify onstraints nd/or oncerns formanufacturing, load paths, and stress concentrations, to develop estimates of structural performance andcosts, and to select a single configuration as showing the greatest potential for success in manufacturing,strength and durability.This report summarizes the work performed on this project, including the approach taken, configurationsand materials considered,and the computational methodology used.The work presented includes:

Candidate fiber orientations and fabric architectures for adaptive blade manufacture are identifiedand assessed on the basisof estimated static strength, stiffness, and fabrication costs.A parametric study is performed for potential blade structural arrangements. Each configuration isevaluated on he basis of estimated manufacturing cost and magnitude of bend-twist couplingachieved.Based on the parametric study results, a single configuration is selected for further evaluation. Acomplete blade model is developed and assessed on the basis f estimated manufacturing cost andbend-twist behavior under steady loading. That design is presently being evaluated via aeroelasticsimulations under the WindPACT Rotor System Design Study.'

Based on heresultsandconclusionsfrom hiswork,recommendationsaremade for follow-onworkconcerning adaptive blades.

1.2 ApproachThe blade design used for this project was selected to be consistent with the WindPACT Blade SystemDesign Study and Rotor System Design Study. The baseline blade assumes a 1.5 MW rotor constructedofall-fiberglassprepreg aminatewithno wistcoupling.Asmany of theadaptivebladesdesignsincorporate a carbon fiberglass hybrid fiber arrangement, a secondary baseline designas developed withall-fiberglass prepreg skins and a carbon fiberglass hybrid spar, again with no bend-twist coupling.Several candidate design concepts were evaluated. Considerations include potential loads reduction, powerperformance,taticndynamictability,tructuralfficiency,manufacturability and estimatedfabrication costs. For each design considered, preliminary structural calculations were performed to sizethe required blade structure. The general process usedwas to:

Develop a standardized loading that represents peak flapwise bending conditions.Estimate the design strains (static compression and tension) for the composite materials proposedConstruct an ANSYS model of a representative blade section using the NuMAD nterfa~e.~Use the ANSYSmodel to determine the schedule of plies required to sustain peak structural loads.Evaluate the resulting mass, stiffness, and extent of flap-twist coupling.Consult with manufacturers to identify candidate fabrication methods, estimate production costs,

using micromechanics and classical lamination theory.

and identify potential difficulties in fabrication and/or obtaining structural robustness.

6


7/24

To evaluate candidate materials and structural layouts, a parametric study was performed using a constantcross-section shape that is representative of blade at the maximum chord location (25% span). Based onthe parametric study results, a complete blade design was developedor a selected structural configuration.

1.3 Concepts InvestigatedTwo basic material types were evaluated for providing the bend-twist coupling. The first is a stitchedhybrid fabric with carbon fibers oriented at a+20 angle (relative to the roll warp direction) nd fiberglassfibers at a second orientation angle. For fabric stability and handling, a small amount of stabilizing fiber isassumed obe ncluded na hirddirection. The secondmaterialconsidered is acarbonpreformconstructed from non-continuous carbon fibers with a high degree f orientation. With these materials, thefollowing structural arrangements were investigated.1.3.1 Biased Skin with Conventional Unbiased Spar CapsIn this approach the skin provides the bend-twist coupling. The unbiased (fiberglass or hybrid carbon /fiberglass) spar caps provide the majority of the flapwise bending strength, but are intended to be relativelycompliant in twisting. The 20 carbon fibers in the biased skin begin and end at the leading and trailingedges of the blade, leading to critical strains at those locations. In a detailed design, it is expected that thecarbon fibers would be curtailed before reaching the trailing and leading edgeso ameliorate this condition.However, for the present evaluation the biased skin is assumedo continue for the full chordwise extent.1.3.2 Conventional Triaxial Skins with Biased Box Spar CapsIn this approach heflapwisebendingstrength and thebend-twistcouplingarebothsuppliedby herelatively thick spar caps, whereas the skins are made from conventional triaxial fiberglass material. Intraditional blade design, the fibersat the upper and lower caps of this spar are primarily uniaxial while theremainingskinsaremadeupofaminimalnumberof riaxialplies.In this typeofdesign, hesparcontributes most of the flapwise stiffness, but it contributes only about 50% of the torsional stiffness.Therefore, making a spar torsionallysoft does not necessarily lead to an overall torsionally oft blade, andit may be necessary to modify the skin layup to achieve the desired amount of bend-twist coupling.1.3.3 Biased Skins and Biased Spar CapsThis design combines the two approaches above, with both the skinsnd the spar caps contributing to thebend-twist coupling.


8/24

2. Candidate Composite Materials2.1 Stitched Hybrid FabricThe nitial fabric architecture considered n this study consisted of carbon fibers at a +20 angle, andfiberglass fibers at -70 angle. For fabric stability and handling, it was assumed that a small amount ofstabilizing fiber would be included inthird direction.GEC worked with Saertex USA LLC to develop specifications and cost estimates for a custom stitchedfabric with this architecture. The initial cost estimates assumed a total fabric areal weight of 923 gramsper square meter (gsm), with 613 gsm of carbon fiber at +22.5 , 300 gsrn of Erglass at -70 , 4 gsm of 0E-glass for stabilization, and 6 gsm of polyester stitching. Taking in account the density of the fibers, thismaterial has 70% carbon content and30% fiberglass content by volume. Based on an assumed productionrate of 325 metric tons per year, Saetrex quoted this fabric at $17.82 / kg, which reflects a relatively highdegree of processing efficiency. In practice, the Saertex stitching process is unable to include non-zeroangles smaller than 22.5 . However, this limitation was confirmed relatively late in this project and anominal orientation angle of 20 has been assumed for the carbon fibers for all the analyses presented.Candidate blade designs were developed using this initial stitched fabric architecture. The initial resultsshowed marginal promise for cost-effective manufacture of adaptive blades, particularly for designs nwhich the biased hybrid fabric is providing the primary structural strength. This can be explained by Table4, which summarizes he modulus and strength estimates made by J. Mandell for a 0 / 90 carbon /fiberglass hybrid laminate at varying rotation angle^.^ The table indicates that while the compressive strainlimits change only modestly with rotation angle, the modulus for supporting flapwise bending loads (Ex)and the corresponding strength drops rapidly.

Table 1. Effect of Rotation Angle on Modulus and Strength (carbonI glass orthotropic laminate)RotationFailurelapending

(Des.) (GP4GP4 @Pa) (X) Mechanismtrength0 69.73.7.87.149.90 Carboniber comp. 1 oo10 42.32.7.14.408.85hearnlassayer.57200.90.7.12.598.91hearnlass

Following hese nitialresults,GEC nvestigatedalternativecandidatearchitectures for stitchedhybridfabrics. Mechanical properties were calculated using spreadsheet analyses that were checked against thecommercial CompositePro code. Some results of his nvestigation are summarized n Table 2. For allcases the carbon fiber contents 75% by volume at a fixed angleof 20 , and the table indicates the strengthand stiffness at varying fiberglass orientation angles. The relative flap bending strength in Table 2 usesthe same baseline properties asTable1(carbon / glass orthotropic laminate at zero rotation). The extent ofeffective bend-twist coupling is indicated by the shear-coupling coefficient, qx,xy,hich is a measure of theamount of shear strain generated in the xy plane per unit strain in the x-direction. Reviewing the failuremechanisms listed in Table 2, it is seen that transverse stress is the governing criterion for nearly everyconfiguration of biaxial fabric considered. The best combination of strength, stiffness, and coupling occurs


9/24

for the fabric with the fiberglass at 0 orientation. However, even for this case the relative flap bendingstrength is only half of that or the unrotated orthotropic fabric.

Table 2. Combinationof 20 Carbon Fiberswith Fiberglass at Varying Orientation Angles

Calculations were also performed to assess the effect of including a small amount of fiber at a thirdorientation angle for the purpose of stabilizing the fabric for handling. When fiberglass properties wereassumed for the stabilizing fibers, the calculations predicted matrix failure in that layer at a strain levelmuchowerhan orheprimarycarbonand iberglassayers.Thisllustrated a generalconceptconcerning the candidate materials under consideration. For the biaxial fabrics listedn Tables 1and 2 , theeffective bend-twist coupling is accompanied by relatively large strains in the matrix. However, stiff fiberssuch as carbon and fiberglass do not accommodate arge ensile strains ransverse to the fibers. As aconsequence the transverse stress limit criterion governed most fabrics considered, even when the ply forwhich first-ply failure was predicted carried virtually no load.It is possible that a more elastic fiber, suchas polyester, could be used n a stabilizing third-axis orientationwithout initiating matrix cracking under load. Another possibility s that a biaxial fabric is preimpregnatedimmediately after stitching, and the B-stage resin could provide stabilization of the fabric during the cuttingand handling required or blade manufacture.Based on the analyses above, two stitched hybrid fabric architectures were selected for evaluation in theparametric analyses of Section 3.0. Both fabrics are 75% carbon and 25% fiberglass by volume, and inboth cases the carbon fibers are at a 20 orientation. The fiberglass orientations considered are -70 andO , with the mechanical properties as shown in Tables and 2 , respectively.2.2 Oriented Non-Continuous Carbon FibersThe second material type considered is a carbon preform constructed from non-continuous carbon fiberswith a high degree of orientation that is under development by the National Composites Center.' Baselinecost estimates and mechanical properties for this P-4A perform in a conventional (unbiased) spar capstructure were estimated during the work of Reference 2 . For evaluation in the present work, GEC usedlaminate heory odevelop hemechanicalproperties of P-4A in a rotatedorientation.Characteristicstrength (strain) values were estimated based on the trends seen for the hybrid fabrics considered above.One potentially favorable attribute of this material is that the orientation angle and perform thickness canbe tailored in both the chordwise and spanwise directions with minimal additional processing costs.


10/24

3. Parametr ic Study of Candidate Structural ConfigurationsA parametricstudywasperformedusing hematerials dentified in heprevioussection nvaryingstructuralarrangements.ForeachcandidateconfiLwration,aconstantcross-sectionblademodelwasdeveloped and analyzed. Each design was evaluated on the basis of estimated weight, manufacturing cost,and magnitude of bend-twist coupling achieved. On the basis of his parametric study, the most promisingcombination of materials and structural arrangement was identifiednd a full-blade model developed.

3.1 Baseline Section Structural DefinitionThebaseline tructuralarrangementwas electedunder heworkofReferences 1 and 2 asbeingrepresentative of current commercial blade designs. The primary structural member is a box-spar, withwebs at 20% and 45% chord and build-up of spar cap material between the webs. The exterior skins andinternal shear webs are both sandwich construction with triaxial fiberglass laminate separated by balsacore. This arrangement is depicted in Figure 1, where airfoil section used for the present calculation (25%span station) is shown. For the 1.5 MW rotor, the blade section has a chord of 2.8 m and a maximumthickness of 0.84 m.

0.30.20.10.0

-0.1-0.2-0.3

balsa-core skins NREL S818 airfo i lscaled to 30% Uc

forward Ishear web

0 0 0.2 0 4 0.6 0 8 1 oXlC

Figure 1. Arrangement of baseline structural modelTable 3lists the layers in the baseline structural shell and describes the material contained in each. Theshear web cores (balsa) are assumed to be 1% of airfoil chord (c) thick, with triaxial skins of 1.27mThe skins and spar cap are E-glass/epoxy laminate. The triaxial fabric is designated CDB340, and has a25%, 25%, and 50% distribution of +45 , -45 , and 0 fibers, respectively. The spar cap is composed ofalternating ayers of triaxialanduniaxial ( A 2 6 0 ) fabric. This stackingsequence esults n spar caplaminate with 70% uniaxial and30% off-axis fibers by weight.


11/24

Table 3. Baseline Structural-Shell DefinitionLaver Thicknessaterial

I 1 I Eel coat I 0.51m2

triaxial fabric0.38 mmandom mat

41.27 mm

O%-15% c28.0 mmalsa5%-85% c28.5 mmpar cap mixture5%-45% c14.0 mmalsa

5 triaxial fabric 1.27mm3.2 Material Properties

Table 5 shows estimates of material and production costs that were developed for the materials considered.These estimates are based on the workf Reference 2 and on subsequent conversations with manufacturersof composite materials nd structures.

Table5 Estimated Cost of Blade LaminateI Estimated Production Cost ($/kg) I

Description Finished BladeaterialFiberglass triaxial skins

20.45 19.20CC P-4A discontinuous oriented carbon$17.6213.12iased carbon fiberglass laminate (spar cap or skins)15.921 1.42nbiased carbon fiberglass hybrid spar cap9.284.58iberglass spar cap laminate9.584.78

Strain-basedvalues for characteristicstrengthwerederivedat MSU for hebaselineE-glass/epoxylaminatesingombinationfestatandaminateollowinghepproachescribednReference2,materialpartial afety actorsweredevelopedbasedonhevalues pecifiedbyheGermanischer Lloyd guidelines.'Table 4 summarizes the values for characteristic and design laminate strength that were used to develop thestructural designs for both the baseline blade and the configurations with bend-twist coupling. With theexceptionofheP-4Adiscontinuous arbonpreform,prepregmaterial ormwas ssumed.Themechanical properties assume fiber volume fractions of vf = 0.45 for all the carbon / fiberglass hybridmaterials, v f = 0 5 for the triaxial fiberglass, andvf = 0.55 for the P-4A discontinuous carbon.

Table 4 Laminate Mechanical Properties andStrength DataIElastic constants (GPa) vxy Density Char. Strength (%) IDesign Strength 46) ITension Comp.

0.370.48 0.480.51 0.37

11


12/24

3.3 Calculation o f Blade Section Structural PropertiesUsing he NuMAD interface,an ANSYS finite-elementmodelwasconstructedforeachconfigurationevaluated. Each model had he general sectional ayout as ndicated n Figure 1. In all cases he shearwebs were balsa core with triaxial fiberglass facings. Each blade design also had balsa-core skins with anouter layer of gel coat and veil mat. The facings of the outer blade section skinsand the spar cap laminatewere modified per each configuration modeled.Each blade design evaluated was modeled within ANSYS as a cantilevered beam of 20 meter length withconstant cross-section dimensions. A tip load was applied, and the strains at the midspan evaluated for a50-year Class 1 extreme gust per the IEC 61400-1 standard.' Based on the results, the thickness of the sparcap aminatewasadjusted terativelyuntil hedesign train evels of Table4were atisfied. Theconvergence criteria used was that the critical laminate strain under load must be within.5 of the designvalue. In all cases, the compressive strain governed the structural designs.Once the blade structure had been sized, the structural propertiesor each configuration were calculated byevaluating the displacementsat the mid-span of a cantilevered beam under unit bending loads. Equation set(9) ofReference 10 was used, introducing the notation:MB = Flapwise bending moment (N-m)MT = Torsional bending moment (N-m)6 Displacementnlapwiseirectionm)

= Torsional displacement due o MT (rads)= Torsional displacement due toMB (rads)

E1 = Flapwisebending sti fhess (N-m2)GJ = Torsionalbendingstiffness(N-m2)

Bend-twistouplingoefficientnon-dimensional)L Length of beam from fixed end tosectionunderevaluation(m)Equation set (9) ofReference 10can then be arranged as:

M , 2 * 6 * $ ,

The overall equivalenr bend-twist coupling :equiv (in units of degrees wist per meter span, per N-m offlapwise bending mom.nt) can be derived by :arranging Equation 3 in the form:


13/24

3.4 Parametric Study Results and DiscussionTable 6 provides a summary of the structural properties and cost estimatesor the configurations modeled.Unless specifically noted otherwise, biased hybrid materials have carbon fibers at an orientation of 20and glass fiber at an orientation of -70 . Although the weightnd cost are evaluatedon a unit span basis, itis expected that similar trends would results at other blade spanwise stations that incorporate bend-twistcoupling. In the following discussion all numerical comparisons will be made relative to the all-fiberglassbaseline (case 1 )The baseline carbon / fiberglass hybrid design with no coupling (case 2) has spar caps that are 35%thinner than the all-fiberglass baseline. Consequently the cost increase for the carbon / fiberglass hybriddesign is only about 1%. Also, the design strain (static compressive) is 0.63Y0for the all-fiberglass baselineand only 0.37% for the unbiased hybrid spar cap. As a result, the hybrid baseline blade would have 35%lower tip deflections under a given bending load than the all-fiberglass baseline. The conclusions that fora conventional blade, the carbon / fiberglass spar cap is competitive with, and perhaps superior to, the all-fiberglass baseline in terms of costnd structural properties.

The relative merit of the seven bend-twist configurations considered is less clear. The configurations havevarying degrees of bend-twist coupling, but show cost increases ranging from 34% to 250% above thebaseline.Forall hecaseswithbiasedsparcaps,significantcost ncreases esult fiom theneed orrelatively thick spar cap laminate. A comparison of cases #3 and 4 shows the benefit of improved loadcarrying capability of biased spar materials. For a spar cap constructed of 20 carbon with -70 fiberglass(case #3), the cost increase over the baseline is 195%. By shifting the fiberglass to 0 (case #4), the costincrease drops o 76%, and he effective coupling is increased 35%. The configuration with fiberglassskins andabiasedsparcapmanufacturedfromP-4Anon-continuouscarbon(case #8) showsacostincrease of 68% over the baseline, but a lower effective coupling value than configuration4.The case with biased skins and a fiberglass spar (case #5) is not promising, with a 90% cost increase andthe lowest effective coupling. This can be attributed to the mismatch in design strains between the biasedskin and fiberglass spar materials. With a design strain of 0.37%, the biased skin material governs thesection design. The fiberglass material has a design strain of 0.63%, but is limited to 0.37% so that theskin design strains are not exceeded. As a result, a substantial amount of fiberglass material is requiredand the spar caps re more than twice the thickness of case l.Of the configurations considered, the best combination of cost and bend-twist coupling is biased hybridskins with an unbiased hybrid spar cap (case #6). The effective coupling is less than case 4,but the costincrease is only 34% over the baseline. The modest cost increase is primarily a function of good matchingbetween the skin and spar cap design strains. The unbiased hybrid spar of case6 is nominally the same asfor the non bend-twist hybrid baseline (case#2).Using the skins rather than the spar cap structure to initiate the bend-twist coupling has additional benefitsconcerning load paths and stress concentrations. It is expected that the biased fibers in the skins could becurtailed prior to the leading and trailing edges to avoid stress / strain discontinuities there. Because theunbiased spars would be carrying the majority of the bending loads, the stress concentrations at the biasedskin drop-offs should be modest. Also, the leading and trailing edges are near the flapwise elastic axis socurtailing the biased fibers in this region should have minimal effect on the coupling achieved. By contrastif a spar cap is providing both bend-twist coupling and flapwise bending strength, it wouldbe much morechallenging to avoid large stress concentrations and matrix-dominated load paths.


14/24

P I


15/24

4. Com plete Adapt ive B lade ModelBasedon heparametricanalysis esults,aconfigurationcomprisedofbiasedhybridskinswithanunbiased hybrid spar cap (case #6) was selected for constructing a structural model for a complete adaptiveblade. This design was developed in conjunction with the WindPACT Rotor System Design Study. Theplanform and twist distribution for the blade are listed in Table 7. The peak bending loads used in theinitial design were based on aeroelastic simulation results from the rotor study (for a carbon fiberglasshybrid blade with no bend-twist coupling). Figure 2shows an ANSYS model of the complete blade. Forreference, the turbine system powernd rotor speed curves are givenn Figure 3.

I f u l l blade AA06 Ir

APB 18 x1 213:41:33NODAL SOLUTIONPLOT NO. 1STEP=ISUB =ITIME=lSIM Y R 4 (RVF)PmerGraphicsAVFES=A11EFACET=IDMX =2 35SMX =.l95EtQ9SMN =-67176

Figure 2. ANSYS Model of Complete Adaptive B lade

Table 7. Planform and Mass I Cost Summary for Initial Adaptive B lade DesignBaselineHybrid lade daptive lade

Spanwise Station Chord Twist Unit Mass Unit Cost Unit Mass Unit CostrlRm) clRm) (Des.) (kglm) ($lm) (kglm) ($lm)0.05 1.75 0.055 1.925 10.4 1074 9,965 1074 9,965

0.07.45 0.055 1.925 10.4 182 1,690 182 1,6900.25.75 0.070.450 10.4 81.8 1,000 78.5 1,3000.50 17.50 0.056.970 2.5 61.6 765 57.3 9500.756.25 0.043.500 0.0 35.5 400 32.5 5301OO 35.00.030.0640.62.515

Totals = 2530g27,625440g32,945Plusootonnection = $32,62537,945


16/24

Table 7lists the mass and cost calculated at each spanwise station for both the baseline (non-adaptivecarbon / fiberglass hybrid) and the adaptive blade. It was assumed that the biased spar material starts atthe 25% span location and that both blades are of identical fiberglass construction between 25% span andthe root. Between 25% and 75% span the unit cost of the adaptive blade is 24% to 33% higher than thebaseline blade. However, a significant portion of the total blade weight and costs located between the rootand 25% span. As a result, the cost for the complete adaptive blade structure is estimated as only 19%higher han hebaselineblade.When $5,000 isadded oeachdesign or he oot-connection, hedifferential for the complete blade assembly is only 6%.Table 8 gives the structural properties for the adaptive blade sections and also tabulates the mean flapwisebending load and resulting blade twist at a schedule of wind speeds. Significant bend-twist coupling isachievedwith hisdesign.At 12 d s windspeed, hemeanchange in wistangle s 3.1 and 4. lo,respectively, at the 75% span station and the blade tip. Above rated, the blade is pitched to feather forpower control and the mean flapwise bending loads decrease substantially. However, the blade will stillrespond in an adaptive way, such that an increase in bending loads will result in additional blade twist tofeather.The blade design presented here, and derivatives of this design, are currently being evaluated via aeroelasticsimulations as part of theWindPACT rotor study. Although these evaluations are not complete, the resultsappear promising for reductions in system loads and cost of energy.

16001400e 1200

5 1000x32 800600tn&' 400200

00 5 10 15 205

Hub Height Wind Speed mls)

40

P20 gv.I8

10 20

ystem Power Rotor SpeedFigure 3. Power and Rotor Speed Curves forAdaptive B lade Model


17/24

I


18/24

5.ConclusionsA number of candidate fiber orientations nd fabric architectures for adaptive blade manufacture have beenidentified and assessed on the basis of estimated static strength, stiffness,nd fabrication costs. The resultsillustrate hedifficulty of achievinggoodbend-twistcharacteristicswhilemaintaining tiffness andstrength.A parametric study was performed to evaluate candidate fabric types and structural arrangements. Theresults indicate:

Configurations hat ncorporateadaptivematerials in he parcapprovide hemosteffectivecoupling. However, the cost premium resulting from the relatively low strength of the adaptivematerials is likely to be prohibitive.Large cost increases were seen for all designs in which there is a significant mismatch between thedesign strain values for the skins nd spar cap laminate.The best combination of effective coupling and cost was achieved for the configuration comprisedof biased hybrid skins with an unbiased hybrid spar cap.

A complete adaptive blade model was developed for this configuration. The blade assembly was estimatedto cost only 16% more than the baseline design. Based on a static bending-load evaluation, a significantamount of bend-twist coupling is achieved with this design. Results from the initial aeroelastic simulationsperformed under the WindPACT Rotor System Design Study indicate that the bend-twist coupling resultsin a meaningful reduction in rotor system loads.

18


19/24

19

6. References1 . Malcolm, D., Hansen C. (June 2001) Results fr om the WindPACT Rotor Design Study. Proceedings

Windpower 2001, American Wind Energy Association, Washington DC.2. Griffin, D.A. (publication pending). Blade System Design Studies Volume I: Composite Technologies

for Large Wind Turbine Blades. Albuquerque, N M : Sandia National Laboratories.3. Laird, D.L. (January 1 1-14 2001). 2001: A Numerical Manufacturing and Design Tool Odyssey.

Proceedings ofA W A S M E Wind Energy Symposium. Reno,N V4. Mandell, J.F. (November 3,2001), Estimated static strength and mechanical properties for hybrid

carbon / fiberglass laminate at varying rotation angles, email communication with GEC.5. Reeve, S., Cordell, T. (May 6-10 2001). Carbon Fiber Evaluation for Directed Fiber Preforms.

Proceedings of the 46th nternational SAMPE Symposium and Exhibition.6. Mandell, J.F., Samborsky, D.D. (1997). DOE/M SU Composite Material F atigue D atabase: Test7. Mandell, J.F. (March 10, 2001), Estimated static strength and mechanical properties for GEC trade-off8. Germanischer Lloyd. (1999). Rules and Regulations IV-Non-Marine Technology, Part l-Wind

Methods, Materials and Analysis. SAND97-3002.study on spar cap materials, email communication with GEC.Energy, Regulation for the Certification of Wind Energy Conversion Systems.

9. International Electrotechnical Commission. (1999). IEC 61400 1: Wind turbine ge nerator systems-Part : Safely Requirements, 2d Edition. International Standard 1400-1.10. Lobitz D.W., Veers, P.S. (January 1998),Aeroelastic Behavior of Twist-Coupled HA WT Blades,

Proceedings of AIAA/ASME Wind Energy Symposium. Reno, V


20/24


21/24

DISTRIBUTIONH. AshleyDept. of Aeronautics andStanford UniversityStanford, CA94305Astronautics Mechanical Engr.

K. BergeyUniversityof OklahomaAero Engineering DepartmentNorman, OK 73069D. BerryTPI Composites Inc.373 Market StreetWarren, RI 02885R. BlakemoreGE Wind13681 Chantico RoadTehachapi, CA93561C. P. ButterfieldNREL16 17Cole BoulevardGolden, CO80401G. BywatersNorthern Power SystemsBox 999Waitsfield,VT 05673J. CadoganOffice of Wind and Hydro TechnologyU.S. Departmentof Energy1000 Independence Avenue SWWashington, DC20585

EE-12

D. CaimsMontana State UniversityMechanical & Industrial Engineering Dept.220 Roberts HallBozeman, MT59717

J. ChapmanOEM Development Corp.840 Summer St.Boston,MA 02127-1533

S.CalvertOffice of Wind and Hydro TechnologyU.S. Departmentof Energy1000 Independence Avenue SWWashington, DC20585

EE-12

Kip CheneyPS Enterprises222 N. El Segundo,#576Palm Springs, CA92262C. Christensen, Vice PresidentGE Wind13681 Chantico RoadTehachapi, CA93561R. N. ClarkUSDAAgricultural Research ServiceP.O. Drawer10Bushland, TX790 12C. CoheeFoam Matrix, Inc.1 123 East Redondo Blvd.Inglewood, CA90302J. CohenPrinceton Economic Research, Inc.1700 Rockville PikeSuite 550Rockville, MD 20852C. ColemanNorthern Power SystemsBox 999Waitsfield, VT05673K. J. DeeringThe Wind Turbine Company515 116thAvenueNENo. 263Bellevue, WA 98004A. J. Eggers, Jr.RAW Inc.744 San Antonio Road, Ste.6Palo Alto, CA 4303D. M. EgglestonDME Engineering1605 W. Tennessee Ave.Midland, TX79701-6083


22/24

P. R. GoldmanDirectorOffice of Wind and Hydro TechnologyU.S. Department of Energy1000 Independence Avenue SWWashington, DC20585

EE-12

D.riffin ( 5 )GEC5729 Lakeview DriveNE Ste. 100Kirkland, WA 98033C. HansenWindward Engineering4661 Holly LaneSalt Lake City, UT 41 17C. HedleyHeadwaters Composites, Inc.PO Box 1073Three Forks, MT59752S.HockWind Energy ProgramNREL1617Cole BoulevardGolden, CO80401D. HodgesGeorgia Institute f Technology270 Ferst DriveAtlanta, GA 30332Bill Holley3731OakbrookPleasanton, CA94588K. JacksonDynamic Design123 C StreetDavis, CA95616E. JacobsenGE Wind13000 Jameson Rd.Tehachapi, CA93561G. JamesUniversity ofHoustonDept. of Mechanical Engineering4800 CalhounHouston, TX77204-4792

M. KramerFoam Matrix, Inc.PO Box6394Malibu CA90264A. LaxsonNREL16 17Cole BoulevardGolden, CO80401S.LockardTPI Composites Inc.373 Market StreetWarren,RI 02885J. Locke, Associate ProfessorWichita State University207 Wallace Hall, Box 4Wichita, KS67620-0044D. MalcolmGEC5729 Lakeview DriveNE Ste. 100Kirkland, WA 98033J. F. MandellMontana State University302 Cableigh HallBozeman, MT 97 17T. McCoyGEC5729 Lakeview Drive NE, Ste. 00Kirkland, WA 98033L. McKittrickMontana State UniversityMechanical & Industrial Engineering Dept.220 Roberts HallBozeman, MT59717P.MiglioreNREL16 17Cole BoulevardGolden, CO80401A. MikhailClipper Windpower Technology, Inc.7985 A r m a s Canyon RoadGoleta, CA93 1 17W. MusialNREL1617Cole BoulevardGolden, CO80401


23/24

NWTC Library 5 )NREL161 7Cole BoulevardGolden, CO80401B. NealUSDAAgricultural Research ServiceP.O. Drawer 10Bushland, TX790 12V. NelsonDepartment of PhysicsWest Texas State UniversityP.O. Box248Canyon,TX 790 16T. OlsenTim Olsen Consulting1428 S Humboldt St.Denver, CO802 10R. Z. Poore, PresidentGlobal Energy Concepts, Inc.5729Lakeview Drive NESuite 100Kirkland, WA 98033R. G. RajagopalanAerospace Engineering DepartmentIowa State University404Town Engineering Bldg.Ames, IA 5001J RichmondMDEC3368Mountain Trail Ave.Newbury Park, CA9320Michael RobinsonNREL16 17ole BoulevardGolden, CO80401D. SanchezU.S. Dept. of EnergyAlbuquerque Operations OfficeP.O. Box 5400Albuquerque,NM 87 185R. SherwinAtlantic OrientPO Box 1097Norwich, VT05055

Brian SmithNREL1617Cole BoulevardGolden, CO80401J. SommerMolded Fiber Glass CompaniesIWest9400Holly RoadAdelanto, CA93201K. StarcherAEIWest Texas State UniversityP.O. Box248Canyon, TX790 16F. S Stoddard79 S . Pleasant St. #2AAmherst, MA 01002A. SwiftUniversity of Texas at El Paso320Kent Ave.El Paso, TX 79922J. ThompsonATK Composite StructuresPO Box160433MSYC14Clearfield, UT 840 16-0433R. W. ThresherNREL16 17ole BoulevardGolden, CO80401S TsaiStanford UniversityAeronautics & AstronauticsDurand Bldg. Room81Stanford, CA94305-4035W. A. VachonW. A. Vachon& AssociatesP.O. Box 149Manchester, MA 0 944C.P. van DamDept of Mech and Aero Eng.Universityof California, DavisOne Shields AvenueDavis, CA 95616-5294

c


24/24

B. VickUSDA, Agricultural Research ServiceP.O. Drawer 10Bushland, TX 79012K. WetzelK. Wetzel & Co., Inc.POBox 4 1534 108 Spring Hill DriveLawrence, KS 66046- 1 153R. E. WilsonMechanical Engineering Dept.Oregon State UniversityCorvallis, OR 9733 1M. ZuteckMDZ Consulting601 Clear Lake RoadClear Lake Shores, TX 77565

M.S. 0557M.S. 0557M.S. 0708M.S. 0708M.S. 0708M.S. 0708M.S. 0708M.S. 0708M.S. 0708M.S. 0708M.S. 0708M.S. 0708M.S. 0708M.S. 0847MS. 0958M.S. 1490M.S. 0612M.S. 0899M.S. 9018

T. J. Baca, 9125T. G Came, 9124H. M. Dodd, 6214 (25)T. D. Ashwill, 6214 (10)D. E. Berg, 6214R. R. Hill, 6214P. L. Jones 6214D. L. Laird, 6214D. W. Lobitz, 6214M. A. Rumsey, 6214H. J. Sutherland, 6214P. S . Veers, 6214J. Zayas, 6214K. E. Metzinger, 9126M. Donnelly, 14172A. M. Lucero, 12660Review & Approval Desk, 96 12For DOE/OSTITechnicalibrary,6162)Central Technical Files, 8945-1

.

Wind Foam Cutters Bades Ect

Documents

Transcript of Wind Foam Cutters Bades Ect