flow3d9.4新功能(E文)

markchan

FLOW-3D News Spring 2009 Development Focus: Improved Sedimentation & Scour Model This Development Note highlights additions to the Sedimentation & Scour model to be released in FLOW-3D version 9.4. The upcoming release of FLOW-3D version 9.4 contains a new and improved sedimentation and scour model. This model replaces the existing sediment scour model in FLOW-3D and contains a number of enhancements: Multiple Sediment Species The ability to simulate up to ten sediment species in one simulation will be possible. Each species can have its own particle diameter, density and sedimentation properties. The existing model in Version 9.3 and earlier versions allowed only a single sediment species, which required users to specify median or “worst-case” sedimentation parameters for simulations. Addition of Scour Potential Model It will be possible to compute the potential for sediment erosion even in simulations where the sedimentation & scour model is not turned on. Therefore, it will be possible to predict where erosion is most likely to occur without the computational expense needed to solve the additional transport of sediment. Additionally, this model can be used as a diagnostic tool with the sedimentation & scour model to visualize the shear stress on packed bed interfaces. Figure 1: Isosurface of packed sediment, colored by shear stress Figure 1 shows a simulation of scouring downstream of a 10 m high weir. The overlying water is not visible, and the packed sediment is composed of three uniformly distributed sediment species, each of sizes 2.5, 6.9 and 10.5 mm diameter and specific gravity of 2.65. Both the sedimentation & scour and scour potential model are used in this simulation—the scour potential model allows users to output the local shear stress at solid walls and packed sediment interfaces. Bed-Load Transport Bed-load transport has been added to the model. In the studies of sediment transport, there are two main modes of motion: suspended-load transport and bed-load transport. Suspended load is what was simulated in the earlier models and predicted that the sediment was entrained at the packed-bed interface and carried into the bulk fluid whereby it is carried downstream. With bed-load transport, the sediment does not become entrained in the bulk fluid, but rather rolls and/or bounces over the packed bed interface. Any size sediment can ** bed-load transport, but generally only finer sediment particles ** suspended-load transport. Therefore, we now have the ability to predict the motion of larger sediment species than before. Figure 2: Two-dimensional slices down the centerline showing the packed sediment volume fraction. Figure 2 shows results from using the bed load transport model in simulations of erosion downstream of a 10 m weir. In the image at left, bed-load transport was not considered. Here, the erosion is presumed to be solely due to entrainment of the sediment into the water. However, because of the presence of gravel (10.5 mm sediment grains), which is not able to be entrained, there is significantly less total erosion than shown in the image on the right, where bed-load transport was considered. Non-Linear Drift Flux Non-linear drift flux is now standard. In the previous model, the drifting of sediment was assumed to follow a linear drag model, where the drag experienced by each particle was a linear function of the relative velocity of the particle to the surrounding fluid. Now a non-linear drag model with a drag coefficient for each sediment species has been implemented. All of these enhancements make FLOW-3D’s new sedimentation and scour model more useful and powerful than ever.

markchan

9.4:
Core Gas, Sedimentation & Scour and new springs and ropes capabilities using the GMO model.

markchan

Development Focus: Springs & Ropes and Sinusoidal Motion— Additions to the General Moving Object Model
This Development Note highlights additions to the General Moving Object model to be released in FLOW-3D version 9.4.

The current General Moving Object (GMO) model allows users to prescribe forces and torques acting on moving objects. In many applications, however, some types of external forces and torques are not known a priori. For example, springs exist extensively in machinery, and ropes are used broadly in offshore mooring systems. Their forces and torques on moving objects are functions of their deformation determined by location and orientation of the moving objects. In the next release of FLOW-3D, users can define springs and ropes that exert forces and torques on the moving objects.


Animation 1: An oil platform fastened to the seabed by 12 ropes.
Click on the image to view the animation.

Animation 2: The same oil platform, but floating freely on the water.
Click on the image to view the animation.

Modeling Multiple Springs & Ropes

The model allows multiple springs and ropes. Both ends of each spring or rope can be attached to any moving object or can be fixed in space. All springs and ropes are assumed to have elastic deformation. A spring can experience compression and/or extension depending on whether it is a compression spring, extension spring, compression and extension spring or torsion spring. A rope, however, only allows extension. At each time step of computation, compression/extension of each spring or rope as well as its corresponding force and torque acting on the moving object are calculated. Inertia of the springs and ropes is not considered, thus a rope’s shape is not calculated when it is relaxed.

Simulating Offshore Platforms

Animations 1 and 2 show an oil platform pushed about by waves under hurricane conditions. In animation 1, a total of twelve wire ropes are anchored at seabed 500 m below sea surface and are fastened tautly on two opposite sides of the platform. In animation 2, no ropes exist—the platform floats freely on the water. In both animations, the waves have a 100 m length and a 10 m height (measured from trough to crest) and are simulated using a fifth-order Stokes wave model (also to be in the next release of FLOW-3D). The wave period corresponding to the wave length is 7.62 s. The deck size of the platform is 90 m by 87 m. Comparison of the two animations clearly demonstrates how the ropes restrict and stabilize the platform’s motion under severe wave conditions.


Animation 3: FLOW-3D users will be able to model sinusoidal
motion, as in this example of a piston pump moving water
into a tank. Click on the image to view the animation. Prescribing Sinusoidal Motion

In addition to the spring and rope capability, users can also prescribe sinusoidal motion to a moving object. Input parameters include amplitude, frequency and initial phase for each sinusoidal velocity component. Animation 3 shows a piston pump moving water from a low level to a water tank. Sinusoidal motion is prescribed to the piston. The opening and closing of the two check valves are driven by fluid flow.

Conclusion

These additions to the GMO model give FLOW-3D users an even greater breadth of applications that they can model. In addition to the examples above, spring forces and torques can be easily applied to valves and other moving mechanical parts in numerical simulations. In ocean engineering, users can study the stability of many kinds of Floating Production, Storage, and Offloading (FPSO) systems. Motions of floating wave power generation systems, mooring of ship and marine measurement equipment and boat towing can also be simulated.

[ 本帖最后由 markchan 于 2009-6-3 09:01 编辑 ]

markchan

Development Focus: Predicting Binder Gas Pressures and Blow Locations in Chemically Bonded Sand Cores
This Development Note highlights the Sand Core Gas model which will be a new feature for FLOW-3D casting users to be released in FLOW-3D version 9.4.

Castings with internal cavities (valve bodies, engine blocks and heads, etc.) can only be made with multi-part mold assemblies. In addition to the drag and the cope, chemically bonded sand cores are “printed” into the mold halves to shape the internal cavities. During the pour of the metal and throughout solidification, the core binders thermally degrade—enough for the core sand to be simply shaken out from the casting cavities.

The insufficient mechanical strength associated with fine core features is typically compensated by chemically bonding core sand. Some of the commonly used binders are polyurethane cold-box (PUCB) and a shell sand binder. The higher than normal heating rates in small cores and the associated higher decomposition rates of the binder are accommodated through the use of specialty sands with high gas permeability and through the use of special venting techniques. For example, in some castings, one drills through the mold to core prints. In others, the core is “shelled” by not letting the binder cure to full depth during the core manufacture.

Monitoring Binder Degradation

Figure 1. Evolution of the binder degradation zone (click the image
to view the animation). A small 2 x 1.12 inches in diameter
PUCB bound core is held in an insulated steel holder and
immersed in iron. (Metal flow field is not shown.) A new model is being developed for FLOW-3D that will allow casting users to monitor binder degradation in cores and realize optimal core venting strategies. The model will predict core pressures, surface locations of binder gas loss into the metal, binder degradation zone evolution and binder gas flow fields. The development is undertaken in collaboration with experimentalists who are providing calibration data for commonly used commercial binders.

Figures 1 and 2 show the results of a binder degradation simulation in one such calibration sample. A small cylindrical, 2 inches by 1.125 inches in diameter core bound with a PUCB binder is held in an insulated steel holder and immersed in iron at 2600 Fº. The binder degrades quickly at the core surface which gives a peak in collected gas volume at about 2 seconds. The complete degradation of the core takes much longer. The binder degradation model captures the timing and the magnitude of the first volume rate peak, suggesting that during the critical part of degradation the model results for peak pressures will be accurate.


Figure 2. Calibration data for PUCB binder and fit with the FLOW-3D core gas model.
The real data can be captured to allow simulations of PUCB degradation in more complex geometries
and under different immersion/flow conditions.
Figure 3a. Core gas pressures and metallostatic pressure in steel. At the top of the core, near the holder, the gas can vent into the metal.

Figure 3b. Mass flux of core gas at core surface. In addition to the desirable venting into the gas collector at the top of the core, the gas is escaping into the steel along the upper side of the core.

To take the modeling one step further, the calibrated model is used to simulate the degradation of binder in a longer, 4 inch cylindrical core immersed at 2900 Fº in steel. Here the data shows that almost half of binder gas is not collected at the top of the core. The simulation predicts (Figs. 3a, 3b, and 4) that for shallow immersion depths a significant amount of core gas is vented into the metal along the upper side of the core. However, at larger immersion depths the core is shown to be sealed by the pressure of the surrounding steel.


Figure 4. Simulated total core surface gas mass flux (black line) and the mass flux into the metal (red line). It is evident that a large portion of binder gas escapes into the metal. At about 40 seconds, the core is sealed again since a sufficient amount of binder degraded and pressure in the core decreased.

[ 本帖最后由 markchan 于 2009-6-3 09:05 编辑 ]

yongxing

太牛了,第一手资料吧.

253029486

顶顶！顶出高手老！

储昭节

9.3不是刚出吗，乍又来9.4，请问各位可有相关的资料学习一下，先谢了。