Screen 1 of 7
The beam reaches the riverbank.
Floating the beam is not the first operation. The stone must reach the vessel before the river journey begins, and between the two lies a bank, a shallow margin the loaded hull cannot use, and a transfer nobody has drawn.
dry ground wet ground shallow water deeper water granite lifting supports. The bank profile, the water depths and the barge outline are drawn dimensions. The beam's 8 m × 1.5 m resting face is the workbook's, as an assumption (Inputs_03!B195, B196).
The load
80 tInputs_03!B187Source design beam mass, resting on a face of 8 m × 1.5 mInputs_03!B195, B196Assumption — 12 m² of contact. The mass is sourced; the face is a chosen value.
The ground
The beam has to cross between bank and vessel twice: once here, and once at the far end. The soft-ground test on the next screen is set at the Giza delivery end, assumed to carry 15,000 kg/m²Inputs_03!B194Assumption. Here at the quarry end the bank is granite bedrock, so that test does not apply — the burden here is the mechanism and the geometry.
The river
The corridor to Giza runs 890 kmRiverHarbor_47!B13Source each way inside a 150 dayRiverHarbor_47!B9Source navigable window. None of it starts until the beam is aboard.
Screen 2 of 7
Resting load versus lifting load.
Lifting does not reduce the beam's weight. It transfers that weight from a broad footprint into much smaller foundations standing on the weakest ground in the sequence.
Where this applies. The Giza delivery end, where the beam comes off the barge onto floodplain ground. The quarry end is granite bedrock, so this test makes no claim about it — ruling MR-03, MethodsRulings_48.
Resting flat
Beam pressure on the ground: 6,666.67 kg/m²GraniteAswan_42!B20Derived, from the 80 t mass over the 8 m × 1.5 mInputs_03!B195, B196Assumption bearing face. Well inside the capacity the model assumes for the bank. The beam itself is not the problem.
GraniteAswan_42!B22
Derived
Raised on supports
Load enters the ground through a lifting footprint of 1.5 m²Inputs_03!B169Assumption, giving 53,333.33 kg/m²GraniteAswan_42!B21Derived. The support outlines and the split of that 1.5 m² into four pads are drawn; the footprint total is the workbook's.
What the bank is assumed to carry, against what the lift asks of it
Bar lengths are in proportion to the figures they carry. Both inputs behind this comparison are assumptions the workbook labels as such: the bank capacity at Inputs_03!B194 and the lifting footprint at Inputs_03!B169. What is derived is the arithmetic between them.
Screen 3 of 7
The missing loading geometry.
The same scene in elevation and in plan, sharing one horizontal axis. Everything the transfer needs has to occupy this space at the same moment.
Vertical and horizontal scales are equal; nothing is exaggerated. Only two dimensions here come from the workbook, and both are assumptions it labels as such: the beam's 8 m × 1.5 m face (Inputs_03!B195, B196) and the 1.5 m² foundation total (Inputs_03!B169). The bank profile, the water depths, the support frames, the pad layout and the barge outline are drawn.
This drawing raises five questions and answers none of them.
- Where does the barge pass? The hull has to reach the beam without crossing the ground the supports stand on.
- Where do the supports stand? Anywhere close enough to reach the vessel is wet ground; anywhere firm enough is too far inland.
- Where is sufficient depth? The water that floats a loaded hull begins offshore of the ground that carries the lift.
- How is the beam kept level? An 80 t block set down out of level loads one corner of the deck, and one corner of the bank.
- How is the vessel held? The hull must stay in position, in a current, while its draft changes by the weight of the beam.
A complete method must place the suspended beam, lifting supports, foundations, vessel, rigging and sufficient water depth into one workable geometry. The workbook does not supply that geometry, and neither does the record.
Screen 4 of 7
Attempt to position the barge.
Move the vessel and watch four requirements trade against each other. Every dimension in this interaction is drawn, not modelled — so its conflicts are shown in amber. Nothing here is a computed failure.
The vessel is sweeping on its own — take the slider to steer it. Drag right to bring the vessel in. The slider continues past the point where the hull is obstructed, so the geometry beyond it can be seen; those positions are not available to a real vessel, and the hull is outlined in amber once it reaches them.
Conflicts are found by intersecting bounding boxes in the drawn geometry. There is no physics here, and no figure on this screen comes from the workbook except the 1.5 m² foundation total.
Each row reports one condition at the current position only. Two of them clear here, and they clear because the vessel is nowhere near the beam. No position in this range clears all four, and one condition clearing is not a feasible operation.
Screen 5 of 7
Loading is not floating.
The river trip still has not begun.
A hull that floats empty at the bank does not float loaded at the bank. Putting the beam aboard pushes the vessel down into water that was already the shallowest on the route.
The two draft lines are drawn, but the hull that carries them is sized to the workbook: its volume to the deck matches the 217.97 m³BargeTimber_43!B12Derived required displacement, and the two lines are the 54.49 tBargeTimber_43!B13Derived hull alone and that hull under the full 101.2 tBargeTimber_43!B9Derived payload. Water depths and hull proportions are drawn.
What the vessel has to float
| Beam | 80 t | BargeTimber_43!B5 |
| Rigging and dunnage | 8 t | BargeTimber_43!B6 |
| Crew | 3.2 t | BargeTimber_43!B7 |
| Provisions | 10 t | BargeTimber_43!B8 |
| Payload subtotal | 101.2 t | BargeTimber_43!B9 |
| Hull timber, worked | 54.49 t | BargeTimber_43!B13 |
| Required displacement, incl. freeboard | 217.97 m³ | BargeTimber_43!B12 |
This page uses the BargeTimber_43 chain, not the 90 t figure at GraniteAswan_42!B27/B28. That is now a formal ruling — MR-04 in MethodsRulings_48, dated 6 August 2026, anchoring all transfer-stability work to BargeTimber_43!B12. The two describe one vessel and differ by a factor of 2.42.
The 90 t basis is not merely a different choice here: it is unusable. The ruling records that on that basis the deck-movement identity used on the next screen understates by 233%, because the unloaded end of the hull lifts clear of the water. That figure is the ruling's own text rather than a model output, so it carries no cell and no status badge. Note also that GraniteAswan_42 rows 27–29 still read the 90 t figure and have not yet been brought into line with MR-04; nothing on this page presents those cells as current.
Why the shallow band matters
The vessel must be close enough to the bank for the beam to reach it, and far enough out to float once the beam is aboard. Those are not the same place. On the drawn profile they are 12.6 m apart, and that distance is not a detail to be settled later: it is the operation.
Nothing on this screen shows a completed transfer. The beam is ashore and the vessel is at the margin, which is where the modelled sequence leaves them.
Screen 6 of 7
The hull will not hold still.
Suppose the vessel could reach the stone. A floating hull is not a quay. The moment 80 tonnes begins to cross onto it, the deck under the load drops away — the whole hull settles, and it also tilts toward the edge the load is arriving at. If the conventional account is right, that movement has to stay small enough for a sledge transfer to survive it.
"Hull in the water" is the waterplane area — the outline of the hull where it cuts the surface. A bigger hull moves less, so this lever is set generous to the conventional account.
The hull is schematic and the up-and-down movement is drawn at 12×, or nothing would be visible at this size. The drop and the tolerance beside it are stretched by the same amount, so the comparison between them is true. The whole hull settles as weight is added, and it tips toward the edge the stone is arriving at; the drop shown is the two together, at that edge.
4.10× over what the transfer can survive GraniteAswan_42!B48Derived
It does not matter how you load it
Pushed on lengthwise, the hull pitches down at the end the stone is crossing. The load sits far from the middle, and the hull's length is what resists the tilt.
The number does not change: on the modelled hull it is 0.410 m either way, and it moves with hull size in exactly the same way for both. Working it out for a tilt along the hull and for a tilt across it gives the same answer, so choosing the other method does not escape it. That non-change is the finding, not a fault in the control.
Both cases come to the same expression — the hull settles by the added weight over the area in the water, and tilts by three times that at the loaded edge. Four times, at the one place the stone is crossing. GraniteAswan_42!B48
And something has to hold the vessel still
Sliding 80 tonnes across a deck takes about 24 tGraniteAswan_42!B53Derived of push. That push has to react against something, and the something is a hull floating free. A vessel at rest in water resists a slow horizontal shove with almost nothing, so the mooring or the grounding takes all of it — every beam, every time, anchored into wet riverbank.
There is no attested provision for it.
Worked at a friction of 0.30Inputs_03!B4Assumption for a wet timber track, which the workbook badges as favourable to the conventional account. The rougher coefficient for unimproved ground would put the figure higher. Ruling MR-05, MethodsRulings_48. The push and the restraint are one number: the reaction equals the push.
What it would take to get inside the tolerance
That comparator is the New Kingdom obelisk barge at 63 × 21 mInputs_03!B200, B201Assumption, giving 1,323 m²GraniteAswan_42!B54Derived and a drop of 0.242 mGraniteAswan_42!B55Derived. It is a thousand years later than the pyramid it is being used to test, and the reconstruction is contested — SourceRegister_01 SRC-BARGE-DIMENSIONS is OPEN at Low confidence. It is here as a sense of scale, never as a ceiling: nobody is claiming it is the largest hull that could have been built, only that it is the largest anyone has evidence for, and it is not big enough.
Both ways out are closed
Let the hull float, and the transfer fails here: the deck moves 4.10× more than the transfer can survive, and 24 tonnes of restraint has to come from a mooring anchored in wet riverbank.
Ground the hull so it cannot move, and the same operation fails on the earlier screens instead: the load now bears through the hull into the ground, and the vessel has to be got off again afterwards — which is NO ATTESTED MECHANISMGraniteAswan_42!B29Policy.
Both routes have been worked through. Neither of them is open.
Screen 7 of 7
What this does and does not claim.
What is claimed
One comparison on this page is decided by the model, at the Giza delivery end. A lifting device standing on the floodplain there puts 53,333.33 kg/m²GraniteAswan_42!B21Derived into ground the model assumes will carry 15,000 kg/m²Inputs_03!B194Assumption — 3.6× over that assumption, and an 8×GraniteAswan_42!B22Derived concentration of a load the ground carries comfortably when the beam simply lies on it. That is the exceedance, and it is the only thing on this page printed in red.
The capacity is an assumption with no source pinned, so the claim is not "it exceeds a measured limit". It is the break-even that carries the argument: the finding holds for any lifting footprint below 5.33 m²SensitivityBreakEven_06!B11Derived, and the model assumes 1.5 m². Because the capacity was taken at the generous end, a lower sourced value would raise that break-even rather than lower it.
What is not claimed
Nothing here reconstructs a working mechanism, and nothing here is offered as evidence of how the Egyptians actually did it. The lifting supports are structural envelopes drawn to occupy space, not proposed devices. Every dimension outside the workbook's own figures exists to test whether the required things can share one piece of ground.
The second finding, NO ATTESTED MECHANISM at GraniteAswan_42!B29Policy, is a statement about the record, not about physics. It stays amber throughout.
The rising flood
The most common answer to shore loading is that the river came to the stone: a seasonal inundation that floats a vessel up to the block, or a canal cut in to meet it. It is a serious answer and this page does not refuse it.
The workbook does not model it, and this page does not price it. It is not refused; it is unpriced.
A flood-assisted method still has to place the beam on the vessel, hold it level, and carry its own cost: the canal or basin and its spoil, the timing against a 150 dayRiverHarbor_47!B9Source navigable window, and the works that make a bank hold a lifting foundation at high water — wetter ground, not drier. Rising water changes where the vessel floats. It does not change what the lift weighs or what the bank can carry.
Scope
The shore transfer is two gates of a five-gate chain in GraniteAswan_42. The other three price different problems and are not quoted here. All five have to close for one beam to arrive.
The diagram does not claim that every conceivable method is impossible. It shows the simultaneous physical requirements that any proposed method must satisfy.