Trim Sheet & Re-ballasting¶

The tank procedure tells you what the glider weighs and displaces today. Most real ballasting work, though, is about change: a new sensor bolted on, a battery or VBD swapped at refurbishment, a fairing upgrade, or simply a mission in water that is heavier or lighter than the last one. This page covers the bookkeeping tool that handles those changes — the trim sheet — and how to move a glider from one target density to another without guessing.

Source

Distilled from several years of service and support correspondence between Seaglider operators, refurbishment teams, and the vehicle's original designers (2016–2024), cross-checked against the Ballasting Procedure and Trim & Flight Model pages. Numbers are worked examples from real gliders, not specifications — your vehicle's own sheet and tank results always win.

The trim sheet¶

Every Seaglider is normally delivered (and returned from service) with a trim sheet: a per-vehicle spreadsheet that lists every part on the glider with its mass, volume, and position along the hull, and from those sums predicts how the vehicle will sit in water of a given density. Layouts vary between versions, but the useful pieces are the same:

Part of the sheet	What it holds	What you use it for
Weight sheet	Every component: quantity, mass, volume, longitudinal position. Includes the variable items — lead strips, foam, nose weights	The single source of truth for what is on the glider
Trim	Totals and predictions for one water density: net buoyancy, internal oil stroke (→ `$C_VBD`), pitch-mass stroke (→ `$C_PITCH`), predicted pitch angle, centers of gravity and buoyancy	Goal-seeking a configuration that is neutral and level
Ballast worksheet	Tank density, measured mass, neutral VBD in the tank, and a "new environment" block: new density + desired thrust → mass change	Turning a tank result into a lead change
Lead worksheet	Where each lead strip and foam piece sits, with weights	The physical install plan
Tank / sea-trial notes, log	Free-text history	Recording what was actually done

Enter data only in the input cells

Most sheets colour-code input cells (typically red text) and compute the rest. If a cell that should be a formula holds a typed number — a removed item still carrying its mass, say — the whole prediction quietly drifts. One sheet reviewed by the designers had a single line stuck at 191 g where the formula should have given 0. Comparing the summed mass with a scale weight of the whole glider is what catches this kind of error.

Keep it true to the vehicle¶

The sheet is only as good as its last update. Anything that changes mass, volume, or position belongs in it before you go near the tank:

New or moved sensors — mass, volume (including brackets and cable assemblies, which are easy to under-estimate), and position. A volume figure that looks suspiciously small compared to similar parts usually is.
Batteries — weigh replacements; primary packs of the same type still differ by tens of grams.
VBD replacement — the hydraulic drive, reservoir, and the oil added to the system. Upgraded VBDs can be noticeably heavier than the ones they replace, so a sheet carried over from the old engine will mispredict.
Fairings — older fairings were solid fibreglass; later ones have a syntactic-foam core with a much lower effective density. Swapping one for the other without updating the sheet can produce errors of hundreds of cc.
Wings and rudder — weigh them. Early on, wing sets were kept with their own vehicle because pairs could differ noticeably; modern sets are closer, but a borrowed set on one glider still differed by ~30 g. Record which set is fitted.
Lead and foam — every piece, by weight and position (see inventory below).

Then weigh the whole glider dry and compare it with the sheet's summed mass. The difference is the sheet's error budget: within a few tens of grams (scale accuracy) is good; hundreds of grams means a missing or stale line somewhere, and it is far cheaper to find on the bench than at sea.

Goal-seeking a configuration¶

With the sheet up to date, the routine is the same whether you are adding a sensor or changing water density:

Set the target density in the Trim tab (see choosing a target density).
Goal-seek net buoyancy = 0 by changing the internal oil stroke. The result is where the VBD will sit at neutral — your predicted $C_VBD.
Goal-seek pitch angle = 0 by changing the pitch-mass stroke. The result is the predicted $C_PITCH.
Check the margins:
- Oil stroke — leaves the thrust you want on both sides of neutral (see thrust margin).
- Pitch-mass stroke — around 70 % is the nominal design point. Much lower or higher eats into the pitch authority you need to dive steeply in one direction.
If a margin is off, move lead, don't just add it. One team fitting a heavier aft sensor first tried compensating with extra lead; the better fix was to move existing lead strips from aft of the joint ring to forward of it, which corrected pitch while keeping total mass (and therefore $C_VBD) where it was. Re-run steps 2–4 after each change.

Adding lead also adds volume

Lead is ~11.3 g/cc, so every gram of lead added also displaces a little water. In seawater, 100 g of lead adds only about 91 g of net weight-in-water. The ballast worksheet accounts for this; back-of-envelope sums often don't.

Calibrating the sheet: the "mystery mass"¶

Sooner or later the sheet and the real vehicle disagree. Classic case: a glider upgraded to a new fairing type had every row of its sheet carefully updated, and the scale weight matched the sheet to within 20 g — yet in the sea it came out roughly 1 kg too light. Foam came off, lead went on, and it flew fine, but the sheet no longer described the vehicle.

The fix is to stop trying to find the error by inspection and instead calibrate the sheet against what the glider actually does:

Take observed centers — the neutral $C_VBD and level $C_PITCH either from a tank trim test or from well-trimmed dives of a real mission — plus the density they were observed at.
Add a line to the weight sheet for a fictitious item (call it "mystery mass" or "ghost volume").
Goal-seek that item's mass/volume and position until the sheet reproduces the observed $C_VBD and $C_PITCH at the observed density.
Now change the density, lead, or sensors you actually want, and read off the new predicted centers.

The mystery line absorbs whatever the sheet doesn't know — fairing density, a mis-measured bracket, a heavier engine — so changes relative to the calibrated state come out right even when the absolute numbers didn't. One team re-used the same calibrated ghost item for a later mission with a different sensor suite and water density, and the glider flew well first time.

Calibrated sheets travel

When a glider goes to a new operator or back from refurbishment, send the calibrated sheet and the observed centers it was fitted to. Without those, the next person starts over.

Re-ballasting for a new density¶

Choosing the target density¶

$C_VBD matters most where the glider is neutral — in practice the density near the bottom of the dive (apogee), and to a lesser degree the surface layer where it has to recover the antenna. Sources, best first:

CTD casts or glider data from the same area and season.
Climatology — Argo float profiles are excellent. One team preparing an open-Atlantic mission pulled ten years of Argo profiles around the track: surface density varied seasonally between about 1025 and 1027 kg/m³, while density at 1000 m sat in a narrow band around 1032. That spread tells you the surface is the uncertain end and the deep value is reliable.
The last mission's numbers — only if the area and season really match.

Write the density down in the sheet and in any work order. Two gliders in one fleet came back from service ballasted for 1027.5 kg/m³ when the operator had requested 1029.2 for a polar mission. The first was only discovered at sea: it sat light and flat at the surface with a poor antenna position, Iridium sessions kept dropping, and the team struggled to trim it through repeated no-comm dives (a damaged antenna cable found after recovery probably made things worse). Reducing $SM_CC helped a little but cannot make up 1.7 density units. The second was caught by reading its trim sheet before deployment.

Check the density cell when a glider comes back

Whenever a glider returns from refurbishment, open the trim sheet and look at the density it was ballasted for before anything else. It takes ten seconds and avoids the scenario above.

How big is a density change?¶

The arithmetic is simple. A Seaglider displaces roughly 52 litres, so

buoyancy change (cc, or g) ≈ glider volume (L) × Δρ (kg/m³)

Density change	Buoyancy change	≈ VBD counts (at ~4 counts/cc)	Lead equivalent (seawater)
0.5 kg/m³	~26 cc	~105	~29 g
1.0 kg/m³	~52 cc	~210	~57 g
1.7 kg/m³ (1027.5 → 1029.2)	~88 cc	~360	~97 g
2.0 kg/m³	~104 cc	~420	~114 g

Going into denser water makes the glider lighter: it either needs more lead, or a higher $C_VBD (neutral reached with more oil inside), which costs you buoyancy margin at the surface. Going into lighter water is the reverse.

Example — absorbing it in $C_VBD. A glider last flown with $C_VBD 2557 at 1027.5 kg/m³ is going somewhere ~1 unit denser at its working depth. 52 cc × 4 counts/cc ≈ +210 counts, so start the mission near $C_VBD ≈ 2770 and let the first dives' regressions refine it. Cross-check with the trim sheet: put in the new density and the stroke the glider actually flew at last time — it should give a very similar answer. If the two disagree, trust neither until you know why. Then check that the remaining thrust is still enough (next section); if not, add lead instead.

Example — rebalancing with lead. A glider ballasted for 1027.5 kg/m³ needs to fly at 1029.2 — about 88 cc lighter in the new water, so roughly 100 g of lead once the lead's own volume is accounted for. Before changing anything, take the fairing off, photograph and weigh every existing lead strip (in the case this is drawn from, five strips of 111–129 g plus eleven foam pieces) and confirm the sheet matches what's on the hull. Then set the new density in the sheet, let it compute the mass change, and split the added lead forward and aft of the joint ring so that the pitch-mass stroke stays near its old value.

Thrust margin¶

Thrust is how much buoyancy the VBD can still produce beyond neutral — it's what drives the climb and what lifts the antenna clear at the surface.

Designers set gliders up for roughly 250–300 cc of maximum thrust in the target water. One refitted glider (two new sensors added, sheet showing ~1.3 kg less mass than before) came out with neutral at ~87 % oil stroke and only ~114 cc of thrust — flagged as "not much" before it left the bench, and a hint that the new sensor volumes in the sheet needed re-checking.
~210 cc is on the light side; operators flying shallow, calm missions can live with it, but in strong currents or rough seas you will want more.
Mission settings like $MAX_BUOY are then a choice within that range — the Trim & Flight Model page covers when to use it.

How much density range can one ballast cover?¶

The VBD's usable range divided by the glider's volume gives the span of water densities one ballast setting can handle. For a ~52 L Seaglider with a few hundred cc of usable stroke on each side of neutral, that's a handful of kg/m³ — enough for most open-ocean missions, but not for plunging from a fresh river plume into full-salinity water, or for crossing a front with a large density jump without re-ballasting. If the mission spans more than about ±3 kg/m³ (leaving some margin), plan to ballast for the critical end and accept less efficiency at the other, or split the mission.

Large foam volumes amplify small density errors

A payload that forces a lot of syntactic foam onto the hull makes the vehicle more sensitive: one heavily foamed glider, ballasted in one sea and deployed in a slightly different one, developed a roll instability that could not be trimmed out at sea. If an integration needs a lot of foam, do the final ballasting in water from the deployment region (or with its density in the tank) rather than relying on the sheet to extrapolate.

Inventory the lead and foam every time¶

Whenever the fairing comes off, record the variable ballast before touching it:

Count and weigh every lead strip and every foam piece. Foam strips are typically ~15 g each, lead bars from ~60 g to ~180 g.
Photograph each face of the hull (bottom, sides, top) with the ballast in place.
Note positions relative to the joint ring / bulkhead — forward or aft matters as much as mass.
Put the numbers in the sheet's lead worksheet and the log.

Why it matters: a glider that behaved differently on two consecutive missions, in a way the water densities couldn't explain, turned out to be consistent with foam having gone missing between them. Without a before/after inventory there was no way to prove it either way.

After launch: when the numbers say the sheet was wrong¶

The first dives are the real ballast test. Signs that the tank or sheet missed:

The glider is clearly heavy or light — the vertical-velocity regression wants a large $C_VBD change, the glider struggles to climb, or sits low at the surface. See Buoyancy trim.
FMS / regression fails to converge with an implausible vbdbias and a suggestion to reprocess with a different volmax. That almost always means the mass or volmax in sg_calib_constants.m is wrong — see When the volume regression blows up.

In both cases, back-calculate: once $C_VBD is well tuned, the known mass and the density at apogee give the true volmax. Put that (and the correct $MASS) on the glider and in sg_calib_constants.m, reprocess the dives, and then feed the observed centers back into the trim sheet as a mystery mass so the next ballast starts from reality.